Savant skills may be widespread in people with autism

[Nils K Oeijord]

http://www.newscientist.com/article/dn16941-savant-skills-may-be-widespr\
ead-in-people-with-autism.html

Savant skills may be widespread in people with autism
* 12:15 14 April 2009 by
Celeste Biever
<http://www.newscientist.com/search?rbauthors=Celeste+Biever>

Savant-like skills, such as astounding memory, perfect pitch or the
ability to multiply very high numbers together, may be much more common
among people with autism than previously thought.

A new study of about 100 adults with autism shows that one third have
skills that stand out, both in comparison with their other abilities and
with the skills of the general population. Previous studies put the
prevalence of savantism in autistic people as around 1 in 10.

"People often focus on the things people with autism can't do," says
Patricia Howlin <http://www.kcl.ac.uk/gsp09/staffinfo/1926> of the
Institute of Psychiatry at King's College in London, who led the study.
"One of the things our study illustrates is that these are people who do
have special skills but they are not being used."

The notion of the savant ? someone who has a skill that is
exceptional both compared to the general population and to that person's
other skills ? has long captured the imagination of cognitive
scientists and the general public alike. But despite this fascination,
the connection between autism and savantism remains mysterious
<http://www.newscientist.com/article/mg20126881.800-inside-the-mind-of-a\
n-autistic-savant.html> .

Some studies indicate that there are more savants within the autistic
population than among the general population and among the populations
of people with other mental difficulties. Putting an exact figure on the
prevalence of such special skills among people with autism, however, has
proved difficult.
Outstanding skills
In an attempt to quantify this, Howlin's team looked at two different
measures of exceptional ability in a group of people with autism ?
all now adults ? who the team have been studying periodically since
they were first diagnosed between 1950 and 1985. They found that 39 met
criteria for either what they call a "savant skill" or an "exceptional
cognitive skill".

To identify savant skills, the researchers sent the parents of the
autistic adults a questionnaire asking them to identify and describe,
using specific examples, any outstanding skills and talents that were
present "at a level that would be unusual even for normal people".

Of almost 100 parents who replied, about half (45) claimed that their
child had a special skill. But only 24 met the researchers' tough
criteria for what constitutes a savant skill: both exceptional in terms
of population norms and above the individual's overall level of ability.

Relying on parents' anecdotal reports of skills could be risky. However,
Francesca Happe
<http://www.iop.kcl.ac.uk/staff/profile/default.aspx?go=10942> , also at
the Institute of Psychiatry but not involved in the study, says that the
team's criteria were "pretty strict": "They didn't count anything that
could conceivably be considered a normal skill. I don't think the
parental reports are inflated."
Searching for savants
Among those skills considered at the savant level were: being able to
name the elevation of both the sun and the moon at any time of day, on
any specified date; being able to name the day of the week for any date
in the distant past or future (a fairly common savant ability known as
calendrical calculation
<http://www.newscientist.com/article/mg19826593.900-are-autistic-savants\
-made-not-born.html> ); perfect pitch; and the ability to say, from a
single chord, which piece of music it came from.

To identify exceptional cognitive ability, Howlin's team also examined
the volunteers' scores on standard intelligence tests consisting of a
range of subtests aimed at different aspects of IQ, such as arithmetic,
spatial and motor skills and memory span. They found that 23 had an
ability on at least one of these subtests that was well above the
general population's average score on that subtest.

Eight of these 23 had also been identified as a mathematical or
calendrical savant according to the first criteria, and the team
concludes that overall 28.5% ? or almost one third ? of their
volunteers had either a savant skill or an exceptional cognitive
ability.

"I think it is a surprisingly high number, but believable," says Happe.
She says that the study opens a window into the mind of a child with
autism and recommends using these isolated, exceptional abilities as a
way to motivate people with autism to learn other skills ? such as
social or communication ones ? that might not come as easily.
One in ten?
Darold Treffert <http://www.daroldtreffert.com/> , a psychiatrist based
in Fond du Lac, Wisconsin who studies savants, says that he sticks by an
earlier estimate of the prevalence of savantism as being one in ten
among people with autism. This is based on previous studies and backed
up by his own observations.

He says this is partly because he is mistrustful of parental reports,
and partly because he does not think that the peak ability in the
intelligence subtests qualifies someone as a savant. "Some autistic
savants do well on IQ subtests, but not all autistic persons who do well
on IQ subtests are savant."

But he says the study is interesting, because it underscores the
failings of IQ tests to measure overall intelligence.

"We are all made up of a series of intelligences, especially the savant,
and IQ measures one component," he says. "Savants starkly challenge our
definition of 'intelligence' and require us to look for ways to measure
other 'intelligences'."

"We need a more reliable definition of savant syndrome, and a more
reliable definition of intelligence," he adds.

Journal reference: Philosophical Transactions of the Royal Society B
<http://rstb.royalsocietypublishing.org/> , DOI: 10.1098/rstb.2008.0328