Welcome to Your Brain (30 page)

sequence is repeated, eventually the change becomes strong enough that the entire sequence can be

triggered by some cue that evokes the beginning of the sequence.

In 1949 the Canadian neuropsychologist Donald Hebb suggested how James’s change might take

place. He proposed that the essential components of learning were the firing of neurons in a precise

order, and the connections between them, synapses, that set up the order. In his formulation, the

strengthening and weakening of synaptic connections between neurons could be the underlying means

by which a sequence of neuronal firing is reinforced. More than twenty years after Hebb made this

suggestion, Terje Lømo and Timothy Bliss proved him right. They found that synapses could indeed

change their strength in a lasting way after being activated (as we discussed in
Chapter 13)
. This

phenomenon, called long-term potentiation, has since been found in a variety of animals, including

primates, rats, rabbits, slugs, insects, birds, and even octopuses. These changes last for minutes to

hours. On longer timescales, connections may rearrange themselves and new ones may grow, perhaps

even leading to structural changes like those seen in the London taxi drivers’ brains.

How do these ideas apply to the hippocampus? Many neurons in the hippocampus excite other

neurons nearby, so one neuron can excite another, which excites the next, and so on—perhaps in long

sequences, all within the hippocampus. This sounds very much like Hebb’s vision of sequences of

activity as a means of reliving an experience. Perhaps the hippocampus’s internal loops of excitation

allow these sequences to be generated.

These loops of excitation might also play a part in why the hippocampus and temporal lobe are so

prone to epilepsy. If these structures have a tendency to form positive feedback loops, then they might

be likely to initiate epileptic seizures, which are periods of runaway brain activity. Indeed, the

cerebral cortex is also full of internal connections—and the cortex is another major site for seizures

to begin.

Chapter 24

Rationality Without Reason: Autism

If you’ve spent much time reading newspapers and magazines over the past few years, you may have

formed the impression that autism is caused by environmental toxins of some sort, perhaps by

vaccination. According to one recent analysis, this idea receives seven times as much attention in the

popular press as it does in the scientific literature on which press accounts are ostensibly based.

Although it makes a good story, the environmental hypothesis does have one major drawback: it’s

most likely wrong—or at least incomplete.

“Autism” is a catchall term for a highly variable set of behavioral disorders that begin in early

childhood. It is defined by three features: lack of social reciprocity, disrupted verbal and nonverbal

communication, and inflexible and repetitive behaviors. Autism affects six out of a thousand people

today and is four times as common in males as in females. People who have normal language but

exhibit the other two features are diagnosed with a related disorder, Asperger’s syndrome.

The social behavior problems caused by autism are very distinctive. One way of describing these

problems is in terms of what researchers call “theory of mind.” This phrase refers to the human

ability to imagine what other people know and what they are thinking or feeling, an ability that

develops in most children around the age of three or four. People with autism have extreme difficulty

imagining anyone else’s point of view, and consequently have trouble recognizing when others are

lying, being sarcastic, mocking them, or taking advantage of them. They have particular trouble with

responding appropriately to faces, including recognizing or remembering them, as well as detecting

facial signals of emotion. Most people pay the most attention to the eyes when looking at a face, but

autistic people tend to look at the mouth or elsewhere in the room.

Sam grew up with an autistic younger sister. As a small child, Karen was late to start talking. As

a toddler, she was prone to hitting other children and shouting at inappropriate times. Talking with

her was an exercise in frustration. She responded to questions such as “How are you?” by repeating

the question, and when prompted to give an appropriate answer (“Karen, say you are fine”), she

replied, “You are fine”—creating endless frustration for both parties. Easily overstimulated, she

spent a lot of time sitting in a corner tapping one finger repeatedly against a finger on the other hand.

This form of self-entertainment seemed to soothe her but was not exactly conducive to group play. As

a boy, Sam didn’t like to have friends over for fear of being interrupted by bizarre yelling or

something worse. He found friends’ houses or the library to be more peaceful than home.

Karen’s problems were apparent enough that she was diagnosed as autistic by the age of five,

which was an early diagnosis in the 1970s, before autism became a well-known disorder. At that time

autism was even less understood by the public than it is now. Her parents spent decades thinking

something had happened to her in early childhood to cause her autism. For example, she was born

prematurely, and they thought her problems might have been caused by rough handling as a newborn,

when the plates of her skull had not fully closed.

A feeling of responsibility or self-blame is common among parents of autistic children; this

feeling has its roots in the assumption that the disorder must have an environmental cause. For many

years, psychiatrists attributed autism to the emotional coldness of “refrigerator mothers”—a complete

misunderstanding, but one that fit well with parents’ feelings of responsibility. In general, diseases

that are not well understood often acquire a reputation of being caused by the environment. Another

example is ulcers, which were long thought to be caused by stress but are in fact caused by bacteria.

We don’t know exactly what causes autism, but we do know that it is a disorder of brain

development with a very strong genetic component. If one of a pair of identical twins has the

disorder, the other twin has a better than 50 percent chance of being autistic, even though twins in

general are not at higher risk for autism than single-born children. Even nonidentical siblings of

autistic children have twenty-five to sixty-seven times more risk of autism than the general

population. And relatives of autistic people have a higher chance of having some autistic symptoms

even if they are not fully autistic.

However, despite the strong contribution from genetics, there is not a single “autism gene.” There

are a few rare syndromes in which autistic symptoms can result from a mutation in just one gene. But

in most cases autism requires some combination of genes to be present. We know this because pairs

of fraternal twins, who share half their genes with one another, have at most a 10 percent chance of

sharing an autism diagnosis. This tells us two things: First, because the environment is likely to make

a similar contribution for both fraternal and identical twin pairs, the effect of environmental causes

must, on average, be weak. Second, the chance that two fraternal twins are both autistic is far lower

than the odds for identical twins. This is a typical pattern of inheritance for a disorder that depends on

multiple genes. To take a simple example, if someone’s autism is caused by inheriting two different

genes containing mutations (let’s say gene A from the mother and gene B from the father), then there is

only one chance in four that the sibling of the autistic person will have exactly the same copies of both

gene A and gene B. For more genes, the chance is even lower. This sort of analysis has led scientists

to conclude that most autism is caused by mutations in two to twenty genes.

Did you know? Monkey see, monkey do: Mirror neurons

Social skills depend on empathy, the awareness of what others are feeling. Empathy is

not present at birth but must be developed in childhood. Studies in psychology suggest that

imitation is one way that children learn to read body language and facial expression in

others. Young children tend to imitate others as if looking in a mirror, moving their left

hand when someone else moves his right hand, and they also tend to imitate the goals of an

action rather than the action itself.

Neuroscientists have found brain circuits that are specialized for imitation and may also

be important for empathy. What researchers call “mirror neurons” are found in the inferior

frontal gyrus and premotor and parietal cortex in monkeys. They are active when the animal

performs a goal-directed action, such as grasping food, or when he watches another animal

perform the same action. Some mirror neurons are active only when the animal sees

someone else make the exact same movement, but others are active when someone else

achieves the same goal in a different way. Some mirror neurons are even activated by a

sensory stimulus that suggests an action that cannot be seen, like the sound of a piece of

food being unwrapped or the sight of a hand disappearing behind a barrier where the

monkey knows there is food. Mirror neurons also seem to distinguish the intention behind a

given action, so that a particular neuron might fire when food is grasped by someone

intending to eat it but not when it is grasped by someone intending to put it away in storage.

These two areas are also active during imitation in human brain imaging studies.

Magnetic stimulation that disrupts the function of the inferior frontal gyrus interferes with

imitation in humans. A major input to the parietal mirror neuron region is an area called the

superior temporal sulcus, which is important for attributing mental states to other people. In

normal ten-year-old children, the mirror neuron areas are more active in individuals with

higher scores on a test of empathy, suggesting that empathy may be learned by imagining

yourself in other people’s shoes.

The social deficits seen in autism may involve a dysfunction in the mirror neuron

system. Autistic children show less activity in these brain areas than normal children when

asked to observe or imitate facial expressions. In addition, the decrease in activity

correlates with the severity of the autistic symptoms. Of course, these findings do not prove

that deficits in the mirror neuron system cause autism, and there are many other brain

regions that do not respond normally in this condition, including the brain area that is

specialized for face recognition. Another possible site for problems in autistic people is the

insula, which is active in processing both one’s own emotional state and that of others (see

Chapter 16
). These promising ideas will attract much more research over the next few

years, which should give scientists more clues about the causes of autism.

Even if autism turns out to be entirely caused by genetic mutations, though, that still leaves open

the possibility that it can be influenced by the environment. A good example of an interaction between

genes and the environment is another disorder, phenylketonuria, which results from a genetic mutation

that disrupts the function of the enzyme that converts the amino acid phenylalanine to another

compound. When phenylalanine builds up in the body, it damages neurons, causing mental retardation

and permanent behavioral deficits. This damage can be prevented by an environmental manipulation

—removing all phenylalanine from the diet.

One argument that seems at first glance to favor an environmental cause for autism is the increase

in diagnosed cases over the past four decades. The numbers seem impressive: there has been a

fifteenfold increase in the reported prevalence of autism since the first studies in the 1960s. On closer

inspection, though, several important factors have changed between early and contemporary studies.

First, the diagnostic criteria are different now, and even a small change in the criteria leads to very

large changes in the measured prevalence. Many kids diagnosed with autism today would not have

qualified when the first criteria were formalized in 1980. Many people who are now diagnosed with

autism would previously have been institutionalized, while others might have been neglected, living