Girls and Autism... | Autism PDD

NY Times Op-Ed
August 8, 2005
The Male Condition
By SIMON BARON-COHEN
Cambridge, England

TWO big scientific debates have attracted a lot of attention over
the past year. One concerns the causes of autism, while the other
addresses differences in scientific aptitude between the sexes. At
the risk of adding fuel to both fires, I submit that these two lines
of inquiry have a great deal in common. By studying the differences
between male and female brains, we can generate significant insights
into the mystery of autism.

So was Lawrence Summers, the president of Harvard, right when he
remarked that women were innately less suited than men to be top-
level scientists? Judging from current research, he was and he
wasn't. It's true that scientists have documented psychological and
physiological differences between male and female brains. But Mr.
Summers was wrong to imply that these differences render any
individual woman less capable than any individual man of becoming a
top-level scientist.

In fact, the differences that show up in brain research reflect
averages, meaning that they emerge only when you study groups of
males and females and compare the two groups' averages on particular
psychological tests or physiological measures. The evidence to date
tells us nothing about individuals - which means that if you are a
woman, there is no evidence to suggest that you could not become a
Nobel laureate in your chosen area of scientific inquiry. A good
scientist is a good scientist regardless of sex.

Nonetheless, with brain scanning, we can discern physiological
differences between the average male and the average female brain.
For example, the average man's cerebrum (the area in the front of
the brain concerned with higher thinking) is 9 percent larger than
the average woman's. Similar, though less distinct, overgrowth is
found in all the lobes of the male brain. On average, men also have
a larger amygdala (an almond shaped structure in the center of the
brain involved in processing fear and emotion), and more nerve
cells. Quite how these differences in size affect function, if at
all, is not yet known.

In women, meanwhile, the connective tissue that allows communication
between the two hemispheres of the brain tends to be thicker,
perhaps facilitating interchange. This may explain why one study
from Yale found that when performing language tasks, women are
likely to activate both hemispheres, whereas males (on average)
activate only the left hemisphere.

Psychological tests also reveal patterns of sex difference. On
average, males finish faster and score higher than females on a test
that requires the taker to visualize an object's appearance after it
is rotated in three dimensions. The same is true for map-reading
tests, and for embedded-figures tests, which ask subjects to find a
component shape hidden within a larger design. Males are over-
represented in the top percentiles on college-level math tests and
tend to score higher on mechanics tests than females do. Females, on
the other hand, average higher scores than males on tests of emotion
recognition, social sensitivity and language ability.

Many of these sex differences are seen in adults, which might lead
to the conclusion that all they reflect are differences in
socialization and experience. But some differences are also seen
extremely early in development, which may suggest that biology also
plays a role. For example, girls tend to talk earlier than boys, and
in the second year of life their vocabularies grow at a faster rate.
One-year-old girls also make more eye contact than boys of their
age.

In my work I have summarized these differences by saying that males
on average have a stronger drive to systemize, and females to
empathize. Systemizing involves identifying the laws that govern how
a system works. Once you know the laws, you can control the system
or predict its behavior. Empathizing, on the other hand, involves
recognizing what another person may be feeling or thinking, and
responding to those feelings with an appropriate emotion of one's
own.

Our research team in Cambridge administered questionnaires on which
men and women could report their level of interest in these two
aspects of the world - one involving systems, the other involving
other people's feelings. Three types of people were revealed through
our study: one for whom empathy is stronger than systemizing (Type E
brains); another for whom systemizing is stronger than empathy (Type
S brains); and a third for whom empathy and systemizing are equally
strong (Type B brains). As one might predict, more women (44
percent) have Type E brains than men (17 percent), while more men
have Type S brains (54 percent) than women (17 percent).

What of Mr. Summers's other claim, that such sex differences are
innate? We know that culture plays a role in the divergence of the
sexes, but so does biology. For example, on the first day of life,
male and female newborns pay attention to different things. On
average, at 24 hours old, more male infants will look at a
mechanical mobile suspended above them, whereas more female infants
will look at a human face.

It has also been found that the amount of prenatal testosterone,
which is produced by the fetus and measurable in the amniotic fluid
in which the baby is bathed in the womb, predicts how sociable a
child will be. The higher the level of prenatal testosterone, the
less eye contact the child will make as a toddler, and the slower
the child will develop language. That is connected to the role of
fetal testosterone in influencing brain development.

Males obviously produce far more prenatal testosterone than females
do, but levels vary considerably even across members of the same
sex. In fact, it may not be your sex per se that determines what
kind of brain you have, but your prenatal hormone levels. From there
it's a short leap to the intriguing idea that a male can have a
typically female brain (if his testosterone levels are low), while a
female can have a typically male brain (if her testosterone levels
are high). That notion fits with the evidence that girls born with
congenital adrenal hyperplasia, who for genetic reasons produce too
much testosterone, are more likely to exhibit "tomboy" behavior than
girls with more ordinary hormone levels.

What does all this have to do with autism? According to what I have
called the "extreme male brain" theory of autism, people with autism
simply match an extreme of the male profile, with a particularly
intense drive to systemize and an unusually low drive to empathize.
When adults with Asperger's syndrome (a subgroup on the autistic
spectrum) took the same questionnaires we gave to non-autistic
adults, they exhibited extreme Type S brains. Psychological tests
reveal a similar pattern.

And this analysis makes sense. It helps explain the social
disability in autism, because empathy difficulties make it harder to
make and maintain relationships with others. It also explains
the "islets of ability" that people with autism display in subjects
like math or music or drawing - all skills that benefit from
systemizing.

People with autism often develop obsessions, which may be nothing
other than very intense systemizing at work. The child might become
obsessed with electrical switches (an electrical system), or train
timetables (a temporal system), or spinning objects (a physical
system), or the names of deep-sea fish (a natural, taxonomic
system). The child with severe autism, who may have additional
learning difficulties and little language ability, might express his
obsessions by bouncing constantly on a trampoline or spinning around
and around, because motion is highly lawful and predictable. Some
children with severe autism line objects up for hours on end. What
used to be dismissed by clinicians as "purposeless, repetitive
behavior" may actually be a sign of a mind that is highly tuned to
systemize.

One needs to be extremely careful in advancing a cause for autism,
because this field is rife with theories that have collapsed under
empirical scrutiny. Nonetheless, my hypothesis is that autism is the
genetic result of "assortative mating" between parents who are both
strong systemizers. Assortative mating is the term we use when like
is attracted to like, and there are four significant reasons to
believe it is happening here.

FIRST, both mothers and fathers of children with autism complete the
embedded figures test faster than men and women in the general
population.

Second, both mothers and fathers of children with autism are more
likely to have fathers who are talented systemizers (engineers, for
example).

Third, when we look at brain activity with magnetic resonance
imaging, males and females on average show different patterns while
performing empathizing or systemizing tasks. But both mothers and
fathers of children with autism show strong male patterns of brain
activity.

Fourth, both mothers and fathers of children with autism score above
average on a questionnaire that measures how many autistic traits an
individual has. These results suggest a genetic cause of autism,
with both parents contributing genes that ultimately relate to a
similar kind of mind: one with an affinity for thinking
systematically.

In order to fully test this theory, we still need to do a lot of
work. The specific genes involved must be identified. It is a theory
that may be controversial and perhaps unpopular among those who
believe that the cause of autism is largely or totally
environmental. But controversy is not a reason not to test it -
systematically, as we might say.

Simon Baron-Cohen is the director of the autism research center at
Cambridge University and the author of "The Essential Difference:
The Truth About the Male and Female Brain."