The Man Who Wasn't There: Investigations into the Strange New Science of the Self (23 page)

The children were first shown the object. When asked what it was, all children answered that it was an egg. Then, the children got to touch the “egg,” and discovered that it was actually made of stone. Once they had the chance to examine the stone-egg carefully, they were asked two questions: “What does it look like?” and “What is it really?” The correct answers would be egg (appearance) and stone (reality), respectively.

About 80 percent of the typically developing children and children with Down syndrome passed such A-R tests. However, only
about 35 percent of the children with autism were able to distinguish appearance from reality: they mainly made the error of saying that the stone-egg looked like an egg and was an egg. “
This suggests that these children alone are unaware of the A-R distinction, and by implication unaware of their own mental states. These results suggest that when perceptual information contradicts one’s own knowledge about the world, the autistic child is unable to separate these, and
the perceptual information overrides other representations of an object
,” wrote Baron-Cohen. The autistic child’s brain is unable to fully utilize prior knowledge to make sense of new information—a clue that the brain’s predictive mechanisms may be impaired.

These studies are telling us something profound. Whatever the brain mechanisms that help us to read the minds of others, the same mechanisms seem to be involved in reading our own mind. How important is reading minds, or having a theory of mind, for our sense of self? “I would consider it crucial,” Uta Frith told me.

But crucial for a certain aspect of our self, said Frith. As we have seen, the self can be broadly divided into two: a prereflective self-awareness (the “I,” or the self-as-subject) and a reflective part (the “me,” or the self-as-object). Theory of mind has to do with the self-as-object. No one is claiming that autism impairs a child’s ability to be a subject of experiences.

If children with autism have an impaired theory of mind—and possibly difficulties reading their own minds—how does this manifest in adults with autism? Do they have trouble introspecting too?

To answer this question, Frith, along with Russell Hurlburt and Frith’s former PhD student Francesca Happé (who has since become
a leading autism and theory-of-mind researcher), tested adults with autism.

Their technique, pioneered by Hurlburt, required the subject to carry a beeper that would beep at random, at which point the subject would have to “
freeze the contents of his awareness” and write down the details of his thoughts. Three high-functioning men diagnosed with Asperger’s, who all spoke and communicated well, took part in the study. The subjects were also given a battery of false-belief tests. What emerged was an interesting correlation between how well they did on those false-belief tests and their ability to introspect.

Two of the subjects, Nelson and Robert, who did well on the false-belief tests, were best at being able to introspect and write down their inner experiences, even though these experiences were reported mainly as visual images and often lacked elements that would make up the inner experiences of normal controls (in normal controls, people whom the autism community calls “neurotypicals,” the experiences included perception of inner speech and feelings).

Peter, the third subject, who did not do well on the false-belief tests, struggled with introspection as we know it. “
There were no reportable inner experience[s] in any of the samples. . . . There were no images, no inner speech, no feelings, etc. that other subjects have reported,” wrote the researchers.

During my conversations with Frith and Happé, I was reminded of James Fahey. I had asked him whether he was close to his sister. He surprised me with his answer. “I’m not saying this is the same for all Aspergers, but for me, I don’t feel [emotions] for people. I don’t have a visceral sense. . . . I don’t get butterflies in my stomach. I don’t get heart palpitations. I love my sister, but it’s done purely at a cognitive level. I
think
love for her; I don’t feel love for her.”

How did his sister react when he told her this? I asked. “She took it rather well,” said James. “I was surprised. I thought that this is going to sound bad to a person who is used to emotional connection with others, that this may come across as quite an affront. [But] she tried to understand. Maybe having grown up with me, she saw that my emotional connection to people was very different from what it was in herself.”

This extraordinary admission reveals multiple things. Firstly, James can introspect, but not about aspects of the self that neurotypicals might take as given. His self-as-object, for instance, does not include sharply felt emotions for others, a deficit that he has overcome by developing an explicit cognitive ability (this is likely why some people with autism who initially fail false-belief tests eventually pass them). But the inability to effortlessly fathom people’s emotions is taxing; he has to compensate by paying conscious attention to, say, body language and facial expressions (something neurotypicals are likely doing on autopilot). No wonder social situations continue to be a source of anxiety. “The computer in my head is working in overdrive,” said James. “It’s stressful. Thirty minutes of socializing can be as draining as a three-hour calculus exam.”

James’s situation made perfect sense to Frith. “It’s just the totally automatic, innate kind of ability that is missing in autism, but that doesn’t mean that they cannot get there in a conscious way, via effort and via learning,” she said.

This feeds into the debate over how an individual develops a theory of mind, the ability to mentalize. There are a few schools of thought. One idea is that theory of mind is a “theory theory”—a confusing phrase that aims to say that we implicitly theorize about what is going on in someone else’s mind, using unconscious cognitive processes. Another
says that we simulate a scenario in our minds in order to understand another’s mind; you sort of go through the motions offline, as it were, of others’ actions and come to some understanding of their state of mind. Yet another view is we can directly perceive another’s state of mind—an inference that happens fast, a process that’s below the threshold of consciousness, and enters consciousness as direct perception.

A deficit of theory of mind might also be linked to another characteristic of people with autism—deficits in so-called executive functions, which leads to difficulties in planning a sequence of actions necessary to achieve a goal.

Susan and Roy recounted Alex’s problems with day-to-day tasks. “A two-year-old child will instinctively know that if you have to go out, there are some things you have to do, like wear shoes and socks, get a jacket on. You automatically know that you need to do these things before you can step out of the door,” said Susan. “For Alex you have to remind him a lot more. Every time it’s a new thing. It’s not something that became habitual with him. Or to know that you have to put your socks on first, then to put your shoes on, or that you have to get your underwear on before your pants.”

Studies have demonstrated strong correlations between difficulties with theory of mind and executive function in children with autism. The argument is that people with autism have difficulty accessing or maybe even representing mental states that model the sequence of possible behaviors that haven’t yet happened (such as putting on socks, shoes, and jacket) or the desired goal (to be fully dressed before leaving the house).

This too supports the notion that theory of mind is not just for accessing the mental states of others, it’s crucial for knowing one’s own mind, and hence for aspects of one’s sense of self. And without insights
gleaned from studying autism, we might never have begun looking for brain structures responsible for theory of mind.

In adults, a set of brain regions is strongly correlated with theory of mind: the temporoparietal junction (TPJ), the precuneus (PC), and the medial prefrontal cortex (MPFC). These brain regions are activated when you think about what others are thinking. MIT’s Rebecca Saxe studied these brain regions in children five to eleven years old, ages when they are developing and honing their theory-of-mind abilities. Turns out the same brain regions are implicated in these children in tasks that require theory of mind. In fact,
the right temporoparietal junction (rTPJ) is most strongly correlated with theory-of-mind abilities in children. Cambridge University’s Michael Lombardo, working with Simon Baron-Cohen and others, has shown that
the rTPJ is functionally specialized for representing mental states, and that this specialization is impaired in people with autism, so the greater the degree of impairment, the greater the difficulties in relating socially to others.

Lombardo and colleagues also implicated another brain region as a problem area in autism:
the ventromedial prefrontal cortex (vMPFC). The study’s subjects (an apt word in this case), both neurotypicals and individuals with autism, were asked questions designed to get them to ponder their own mental characteristics or those of the British queen: “How likely are
you
to think that keeping a diary is important?” versus “How likely is the
queen
to think that keeping a diary is important?” The neurotypical controls used the vMPFC more on tasks that referred to the self than to the queen. This was not the case for people with autism: “the ventromedial prefrontal cortex treated self and other equivalently in autism.”

The same study found another intriguing link: with autism, there was reduced connectivity between the vMPFC and regions of the brain that are involved in more basic body representations, including the ventral premotor and somatosensory cortex. Couple this with the fact that the right TPJ (the other region strongly implicated in the theory-of-mind deficit, as we’ve seen) is also linked to body maps, and a new way of looking at autism begins to emerge. It’s a perspective that some researchers are taking seriously: that autism may be due to an inability to accurately perceive one’s own body and the sensory stimuli that it’s receiving. This would disrupt one’s sense of the bodily self, and while such an interruption would have direct consequences for processing sensations, it could also affect higher-level processes such as theory of mind.

There’s an incident from Alex’s first day in first grade that Roy cannot forget. Alex was sitting at a desk with a girl he knew from day care. The teacher gave Alex and the girl some crayons and paper and asked them to draw. The young girl drew a beautiful butterfly. “Really, really cool butterfly,” said Roy. But Alex barely managed some “random stuff,” said Roy. “It was way behind the standard which you expect from a first grader.”

Roy wasn’t just being a harsh, competitive parent. Alex had genuine difficulties with drawing, both in terms of having fine motor control over the use of pencils or crayons and in being able to draw meaningfully. When he’d use a pencil, even to write, he’d not press down hard enough on the paper. Over the years, Alex’s occupational therapist spent hours teaching him how to draw a body. The most Alex could manage were stick figures, even though the therapist explained
that the neck was not a stick, it had some breadth, or that hands were not just lines that ended in points, they were made of palms and fingers. “His drawings were so primitive,” said Susan. And even today, many years later, ask Alex to draw a human shape, and the hands will most likely end in a circle with five smaller circles for fingers. “He still hasn’t got it,” said Roy.