Dr. Laurent Mottron is a professor of psychiatry at the University of Montreal, and holds the Marcel & Rolande Gosselin Research Chair in Cognitive Neuroscience of Autism at the same university. In an email, he explained to R. Prasad the qualities that people with autism have, and the importance of measuring their actual intelligence using non-verbal tests. Excerpts:
For how long has the research assistant Michelle Dawson been working in your lab?
For the past seven years.
Is the learning skill seen in Ms. Dawson unique to her? How prevalent is it in autistics?
She is both autistic and exceptionally intelligent, which is not the case with all autistics. From non-autistics' perspective, we say she is not intelligent because she is autistic, but it is her autism that gives her intelligence a special quality. The laws we detect in autistic cognition are found at the group level in most autistics. When you average the performance of autistics and then compare their performance with non-autistic people, you will find the superiorities I have described in the Nature paper.
At the individual level you may not find the superior abilities in all individuals, but at the group level they have them. However, some of their strengths (the visuo-spatial ones) are not found in people with Aspergers, who have strengths of another kind, mostly language-related. Moreover, autistic abilities are modulated by their intelligence level: autistics, like non-autistics, have a Gaussian distribution of intelligence. But even autistics with moderate intelligence can be very good in their ‘special' abilities.
Do you see the skill at comparable levels in the other seven people working in your lab?
No. She is outstanding in this regard. Others have the same type of qualities, but to a lesser extent.
What is the currently used method to measure the various skills in people with autism?
Classical neuropsychological tests assessing a representative sample of human cognitive abilities are generally used. People also sometimes use a non-verbal test like the one we recommend — Raven Matrices, but success at this test is usually not recognised as a measure of actual intelligence.
Which non-verbal tests do you think should be used for these people?
It should be Raven matrices, and Block design for younger people.
Do scientists routinely use these non-verbal tests?
No, not routinely. Non-verbal tests are used only in about one in seven autistics, and mostly in research.
If only about 10 per cent of autistics have intellectual disability, why has it been estimated to be about 75 per cent for a long time?
We say that among autistics, around 10 per cent have another condition, mostly genetic or neurological or both, which frequently affects intelligence and produces a “true” mental disability. We say that in autistics without this type of accompanying condition, mental retardation has to be measured, not simply inferred from the failure in verbal tests.
We do not know the actual proportion. According to our sample the proportion of those with actual mental disability should be much lower than what is currently considered.
The problem is confounding intelligence and adaptation. This confusion is not seen in the case of other disabilities (see for example the recognition of S. Dawkin's intelligence), only in autism. Autistics have a strong adaptation problem. But their adaptation problem has nothing to do with intelligence.
Has the scientific community ignored the autistics despite knowing their many strengths?
Most scientific studies have put the emphasis on social impairment in autism and consider their strengths with a kind of curiosity which is both poorly relevant at the scientific level and has no adaptive value. We disagree with that. Some researchers, like U Frith and F Happé in U.K, have emphasized that autistics were better at some things than others, and labelled this as “islets of abilities.” Others like P. Heaton and K. Plaisted have also found cognitive superiorities. We worked in continuity of their work, found a lot of other superiorities, which led us to modify classical interpretation of autistics' “islets of abilities.” We now consider that “islets of abilities” represent actual intelligence, rather than exceptions, or preserved cognitive function among a large number of impairments.
If these people have unique capabilities complementing those in non-autistics, should we not involve more of them in research?
Yes, but not only in research but in working places in general.
Do you think that they will excel only in certain areas of research?
I am not sure we can restrain that to the clichés of autistics being good only at manipulating data. As soon as you have people to take some care of their needs, it may work. Often they are good only in a part of what represents a ‘job.' This is one of the difficulties in including them in work settings.
What special training should autistic kids undergo to be absorbed into research?
I am not sure if it has to do with training. According to Michelle Dawson, they should have access to a large set of data (computerised or not) as soon as they display an interest in it.
Since they have communication problems, can they work as a team in a lab with non-autistics?
This needs some adaptation. Some aspects are quite difficult with autistics like changing plans, deadlines, and levels of detail. Somebody should be there to mediate their contribution often, not always.
Despite reduced interpersonal relationship, can they work as a team with non-autistics? Or should you group all the autistics as one team? What has been your experience?
Important question. We do not advocate segregation, but for tolerance and acceptance by non-autistics. When you dine every day with an autistic, you just must accept that his conversation will not follow the same rules as yours, and be prepared for it. They may also struggle with some aspects (productivity, competition) from which they should be protected.