point to.
Consider these examples, all from Jessy’s third decade. «I better remind Daddy about my dental appointment». «We are having chicken livers for dinner». «My supervisor said I’m going away for Christmas». The first two seem to be perfectly ordinary statements, and the third, though a little odd, is at least possible. Except that it wasn’t Jessy’s appointment, it was her father’s. Except that she’d already had dinner; she wasn’t included in that «we». Except that it was the supervisor, not Jessy, who was going away. Jessy knew all these things; she is the last person to be confused about matters of fact. It was the words, the slippery, shifting words, she couldn’t handle.
Kanner had it right. She couldn’t change the pronoun «to suit the altered situation», least of all when she had to manage two pronouns and correctly relate them to each other. Pronouns shift. Nouns stay still. She had no trouble with «Miranda», as she had spoken her own name years before she referred to herself as «I».
The altered situation: that’s the key. And what has altered the situation? Its external constituents are unchanged: the dental appointment, the chicken livers, the vacation, even the people referred to. What has changed is the interior situation, the point of view. Point of view is what determines the choice of pronoun. His appointment. You — you two — are having chicken livers. She’s going away. Pronouns must be adjusted, changed from first to second or third person, from singular to plural, often both at the same time. It’s complicated even to write about; how do little children manage to learn to do it?
Recent research suggests that they do it, easily and naturally, because they have a «theory of mind». Never mind the grandeur of the phrase; consider instead the implications of the experiment Uta Frith and her colleagues reported in 1985. The situation is simple: an experimenter, a child, and two puppets, Sally and Anne. While the child watches, the puppets act out a little drama. Dr. Frith describes it: «Sally has a basket, Anne has a box. Sally puts a marble into her basket. Sally goes out for a walk. While Sally is out of the room, Anne (naughty Anne!) takes the marble from the basket and puts it into her own box. Now it is time for Sally to come back. Sally wants to play with her marble». Now the investigator is ready to ask the key questions: «Where will Sally think her marble is? Where will she look?»[12]
Obvious, isn’t it? A normal four-year-old will get it right: Sally will think it’s in her basket, because that’s where she left it. It’s equally obvious to an autistic child. Sally will look in the box. Why? Because that’s where the marble is; he saw Anne put it there. He doesn’t consider what Sally thinks. He doesn’t know what she thinks. Autistic children, writes Frith, «have no problems in understanding what it means to see and not to see something.. but they cannot understand… somebody else’s attitude or belief».[13] They have no «theory of mind» — of what goes on in other minds. Without that, how can they make sense of how people speak of others, how they speak of themselves? A normal two-year-old may reverse «I» and «you». But it isn’t long before he works it out spontaneously. He hears his mother say «I» and realizes she means herself; soon he reflects that realization in his own speech. He’s recognized two perspectives, hers and his own, and adjusts accordingly. He doesn’t need to echo pronouns; he understands them.
Of course if a child cannot recognize attitudes and beliefs, the effects reach far beyond language — beyond mere grammar or fluency into all the interactions, spoken or unspoken, between human beings. Frith’s subsequent research confirms what parents know and autistic adults themselves report: that even if speech at length develops normally, even if I.Q. tests as superior, the difficulty in perspective-taking remains. That social handicap is at the core of autism, and it will occupy the fourth part of this book.
Pronouns are only a few of the words whose choice depends on our perspective, the view from where we stand. «Here» for you is «there» for me; «ask» for you is «tell» for me. Jessy reverses these too. In «I am my niece», it isn’t only the pronoun that is skewed. «She is»: the verb must shift with the person. There are all sorts of words that give Jessy trouble. Consider «some» and «any», and their paired but very different meanings. When Jessy remarks, as she did one day at lunch, «Sometimes people don’t like to do anything», it sounds like an observation on the human tendency to laziness. Still, I’m puzzled; this remark seems to have no relation to what we’ve been talking about. But when she adds that «people are different» I understand. Change «anything» to «something» and it comes clear. Jessy wanted Dad to balance our checkbook right away, and he said he didn’t feel like it. (People are indeed different; Jessy loves to balance her checkbook.) The problem is long-standing; I recall the eight-year-old who converted «somebody» and «nobody» to «one-body» and «zero-body», trying vainly to pin down what is indefinite by its very nature.
Such peculiarities are only partly explained by a difficulty in perspective-taking. They point us to another core characteristic of autism: to that «anxiously obsessive desire for the maintenance of sameness» which Kanner noted in 1943.[14] Change happens, of course; it can’t be avoided — not outside Nirvana. But tell that to the child who’s shrieking because the bus has taken an alternative route to school, or because a random act has disrupted a pattern only she could see. It can’t be avoided in the world, and it can’t be avoided in that reflection of the world that is human language. From the beginning of its slow development, Jessy’s language was pointing us to the cognitive disability hidden in those emotional reactions. She wanted the world to stay still, needed it to stay still, because if it would only stay still she could understand it.
The children relate better to objects than people. Kanner noticed that too. Objects don’t change. People do; their expressions, their voices, their every word. Jessy understood objects. When she was still speechless and uncomprehending, she stacked blocks, put rings on a stick, sorted shapes and colors. At three, she surprised the psychologist who tested her by rapidly fitting twelve different shapes into a form-board. He concluded she had no mental deficiency. And she didn’t — not in the realm of the unchanging, the absolute, the thing that is what it is. (I’ve written at length about this in The Siege, so I’ll be brief here.) It was relative concepts — and the words that express them — that divided what she could learn easily from what she couldn’t master. As I wrote in that earlier book, «What she was able to grasp were absolute terms… — those that reflected concepts that could be. understood in themselves. ‘Box, ’ ‘cat, ’ ‘giraffe.’ ‘Rectangle, ’ ‘number, ’ ‘letter.’ What she could not understand were relational terms — those that must absorb their full meaning from the situations in which they occur».
A giraffe is a giraffe wherever you find it; a rectangle is a rectangle. Not so with nouns like «teacher», «friend», «sister». My teacher may be your sister, or her friend. It was not until late in Jessy’s teens that we could teach her, with charts and written examples, the simple words for generational relationships. Or are they simple? Now, at forty, she can say Miranda is her niece, getting the noun right if not the verb or pronoun. She knows she herself is Miranda’s aunt and my daughter. If I asked her, I think she could tell me that her sisters are also Miranda’s aunts, that her brother is Miranda’s father, that her mother is Miranda’s grandmother and her niece is my granddaughter. Even now, though, I’m not sure she could manage more than one of these shifting relationships in the same sentence.
Certainly this too is a matter of perspective-taking, of «theory of mind». (House for you is home for me.) But we must ask a further question: Why is it so hard for autistic children — and adults — to assume another’s point of view? In Jessy’s childhood I was content to think vaguely of a «social instinct» that normal children had and Jessy lacked. At that time there had been little research into the way these children’s minds actually worked; Sally and Anne were far in the future. There had been even less interest in the brain pathology that might underlie autistic characteristics — not surprising when psychiatric opinion was generally satisfied that autism resulted from damage done to a previously normal child unlucky enough to have a refrigerator mother.
In the past twenty years, however, biological research has taken off. In 2000 alone the National Alliance for Autism Research, founded (by two parents) only five years earlier, was able to fund sixteen new projects. Many others are sponsored by universities, medical schools, and the National Institutes of Mental Health. (Informed and useful summaries of promising work may be found in NAARRATIVE, the alliance’s quarterly newsletter. See Appendix III.) They lead in many different directions — into the mechanisms of information processing, the possibility of a mutated gene, the role of serotonin in brain development. Particularly significant in understanding what might explain that missing «social instinct» is the work of Eric Courchesne, professor of neuroscience at the medical school of the University of California at San Diego.
Courchesne’s research allows us to make sense of so much that we’ve observed in Jessy: her problems with pronouns, with relational language, with perspective-taking, with people, with everything that requires her to respond flexibly and rapidly to change. His hypothesis — and through magnetic resonance imaging he has amassed hard evidence for it — is that damage to particular locations in a baby’s cerebellum reduces the capacity to shift attention from one sensory stimulus to another. Consider an experiment in which the subject must disengage his attention from a visual cue — a red flash — and shift it to an auditory cue — a high tone. Autistic children take up to ten times longer to do this than normal, or even retarded, controls. Even high-functioning adults take three times
