Autism is a developmental disorder of the brain. It exists from birth and persists throughout life. The causes of the disorder are still unknown, but a strong genetic component as well as other organic causes have been implicated.
Asperger Syndrome is a variant of autism and yet it has distinctive features, in particular, good verbal intelligence, which make its clinical manifestation quite different from classical autism. It is just as much a brain based disorder and again there is a strong genetic component.
For the rational design of treatment programmes, and indeed for the proper understanding of the disorder, it is essential to have a full account of the brain basis of autism and Asperger syndrome and their manifestation in behaviour across all stages of development. We are still far removed from this desirable state of knowledge.
However, cognitive explanations of autism have made considerable progress. In particular, the 'theory of mind' account has vastly increased our understanding of many of the puzzling social and communication impairments of autism. But what is meant by the term 'cognitive'? The cognitive level of explanation refers to what lies between brain and behaviour: the mind. Put simply, the brain produces the mind, and the mind produces behaviour. So when we talk of cognitive processes we are talking about mental processes. Mental processes can go wrong in brain based disorders. How is it possible that, when there is an abnormality of neurological development, some of these processes fail while others run perfectly well ? This question is central to cognitive explanations.
Behaviour suggestive of autism can already be discerned in infancy, but with increasing age the tell-tale signs can become camouflaged, especially in cases with fluent language and without general intellectual delay, as in Asperger syndrome. Asperger syndrome is still a controversial diagnostic category, but it is clear that there are different degrees of autistic impairments and the milder forms can be quite well distinguished when observing an individuals developmental course. Asperger syndrome seems to be a 'milder' variant because people affected by this syndrome have a certain amount of social insight which allows them to understand the minds of other people.
The notion of a whole spectrum of autistic disorders of different type and severity, not confined to only 'nuclear' cases, as originally describes by Kanner, is now well accepted. One of the major problems with behaviourally defined developmental disorders is how to identify and agree on the core features. During development behaviour is extremely changeable. It varies not only with age and ability, but is modified by a multitude of environmental factors. This problem is compounded by the fact that many affected individuals show additional and possibly even unrelated problems, as well as secondary problems. The search for a common denominator can only be carried out at die cognitive level of description. When we only look at behaviour, then it is almost impossible to find a common denominator.
It was known that autistic children, compared to other handicapped children, showed extremely poor pretend play. Hence the hypothesis that autistic children lacked Leslie's critical mechanism suggested itself. The hypothesis that autistic children cannot keep track of mental states separately from physical states was originally tested by Baron-Cohen, Leslie & Frith (1985). The original test was extremely simple and is worth examining here in some detail, because it Illustrates what it means to have an everyday 'theory of mind'.
Normally developing children of about 4 years can reliably answer this question, and so, perhaps surprisingly, can children who are mentally retarded due to Down syndrome. They say that Sally will look for her marble in her basket where she had herself put it. When asked, they can also indicate that Sally did not know what Ann had done (since she was out when Ann took out the marble).
By showing this understanding young children demonstrate that they have attributed a mental state. They grasp that someone can have a false belief about a situation. The 'false belief' is a mental state that is different from a physical state. What good does this do to the child? The mental state can very helpfully explain and predict someone's behaviour. For instance, the false belief predicts that Sally will look for her marble in her basket - and that she will not find it there. Understanding false belief naturally goes together with understanding true belief, knowledge and ignorance. It also goes together with understanding intentions, desires and feelings. This is what is meant by having a 'theory of mind'.
Most children with autism, even of a mental age far in excess of four years, find the simple Sally Ann task a great puzzle, and they tend to get it wrong. They say that Sally will look for the marble in Ann's box, (where it really is) - even though they remember correctly that Sally had put the marble into her own basket and that she was not present when Ann transferred it to her own box. Despite watching and remembering the simple sequence of events, they cannot make sense of the events by inferring that Sally must have a false belief.
They do not take into account at all what Sally thinks: They miss the important change in her mental state (her previously correct belief is now wrong). Thus they cannot predict Sally's behaviour. Clearly, they would get very confused in such a situation in real life where they would find it odd that Sally looked for her marble in the wrong place. Their lack of understanding of false belief goes together with a lack of understanding of other mental states. Hence the claim that autistic individuals do not have a 'theory of mind'.
In a study with volunteers with Asperger syndrome this same group of researchers found that Brodmann's area 8 (left medial frontal) was significantly less active than in the normal group. Such a result is suggestive of a localised and specific brain dysfunction. A dysfunction in this circumscribed part of the left frontal cortex might have a variety of causes. It is clear that more investigators are now warranted with the hope of pinpointing a critical brain area involved in the core deficits of autism. Guided by the hypothesis of a particular cognitive deficit, anatomical studies of the autistic brain can become more targeted.
In fact, on closer inspection, Asperger syndrome individuals who pass theory of mind skills in the laboratory with flying colours, still have problems in applying their theory of mind skills in real life. First of all, they tend to be in their teens and older and have relatively high verbal IQs . No child with autism has yet been found who demonstrated understanding of false belief at the right age, i.e., four to five years. Furthermore, their ability to mind-read seems to remain blunt and slow. Thus, it seems likely that they have achieved an understanding of mental states by a different and unusual route. For this reason, there is a different quality to their everyday mind-reading. The first idea of lack of theory of mind as a common denominator therefore has to be modified: there is mindblindness, but there is also mind-myopia. And again we may come to see that there are different degrees of mind-myopia.
Despite its success in explaining the triad of impairments the theory of mind deficit account has severe limitations. It does not address the reasons for the varying degree of severity of symptoms in autism. It does not address the question why mental retardation is so prevalent in autism. Even in relation to the specific features of social impairment in autism much still needs to be explained. For instance, the hypothesis says little about the often reported poor emotional responsivity.
The main limitation of the theory of mind hypothesis is perhaps that it does not address a number of highly prominent persistent and handicapping features which are increasingly recognised as in need of explanation. 'Restriction of interests' and presence of 'stereotyped' and 'perseverative' behaviour are examples. Even less explained are the islets of ability and special talents which are present in a sizeable proportion of autistic individuals, and more generally, the characteristically spiky profile of IQ test performance. These features seem to apply just as much to individuals with Asperger syndrome as to individuals with autism. That is success or failure on theory of mind tasks is quite independent of the presence and severity of the special features mentioned above. Two cognitive theories currently tackle these unexplored features.
Some of the puzzling features in everyday behaviour of individuals with autism, such as rigidity and perseveration, are also found in patients with frontal lobe damage. By analogy, this suggests the existence of frontal lobe impairment. On the other hand, it is striking that patients with frontal lobe lesions do not have autism. One possible explanation of this fact is that brain damage from birth and brain damage in later life present a very different picture, even if the same brain area is affected. A brain abnormality early in life would be expected to have wide-ranging effects on many aspects of development, while a later acquired lesion may have more specific consequences.
The theory of autism as a disorder manifested particularly in executive functions has much to be said in its favour. It does not replace the theory of mind deficit account, but has the potential to explain a number of features of autism that are not tackled by this account.
Unlike normal people, autistic individuals do not seem to be disadvantaged when processing meaningless, or unconnected material compared to meaningful and connected information. In the normal case there tends to a huge advantage in favour of meaningful material. For instance, meaningful sentences making up a story are much better recalled than jumbled sentences.
In the visuo-spatial domain the advantage of autistic individuals on the Block Design subtest of the Wechsler scales may be due to the spontaneous mental segmentation of the designs into unconnected and 'meaningless' units which happen to be highly appropriate for the reconstruction of the overall design with the given blocks.
In a language task able people with autism read English homophones (e.g. 'Lead', related to leading vs. lead, the metal) - separately and in context. Weak ventral coherence would predict a relative independence from context, and therefore we expected telltale errors in the pronunciation of the homophones, e.g., reading lead as in the metal, when in fact the context demanded the meaning of lead as in leading somebody. This was indeed found. There is thus some evidence from verbal and nonverbal tasks for a distinctive cognitive style in autism.
The drive for coherence appears to be an organising principle in normal human information processing. If this drive was weak in autism, then certain otherwise puzzling aspects of the intelligence of autistic individuals might be explained.
These paradigms are perfectly suited to the subtraction technique that is used in functional brain imaging experiments. We hope therefore that future studies of brain activity during the performance of critical tasks, including those of executive functions and central coherence will yield important clues to brain abnormalities.
After the initial enthusiasm for a single cognitive deficit that could best explain the core features of autism, research efforts have widened considerably. It seems unlikely that one single cognitive abnormality can be identified that would explain all the abnormalities present in autism. The existence of multiple deficits at the cognitive level is a realistic alternative and might help us understand why autism can exist in many different forms ranging from mild to severe. The explanation of autism at the behavioural and cognitive level needs to be complemented by the explanation at the biological level. Cognitive theories are well suited to help in this endeavour and have already started to guide the search for the brain basis of autism.
Uta Frith
MRC Cognitive Development Unit
4 Taviton Street
London WCIH OBT
and University College London