So natural selection can ask, “What kind of information, available to a child, is the best basis for forming long-term patterns of behavior?”, run the experiment, and come up with the answer. Natural selection can ask, “Which is better, to have children’s personalities shaped by their experiences within the family or by their experiences outside the family?”, run the experiment, and come up with the answer. The mental organ with the design specs that enabled it to outperform the other candidates for the job was one that wouldn’t be hijacked by random events or misleading experiences, but would collect as much relevant data as it could, for as long a period as it could afford to, because ugly ducklings sometimes do turn into swans. Being favored by a parent tells a child nothing that will be useful to her later on, but being given special treatment by many different people is informative. Being dominated by an older sibling conveys no useful information to a child, but being dominated by children of his own age is informative. Being beaten up once by a bully is a random event; being pushed around for years by many different adversaries is informative.
In the trial-and-error experiment that produced our species, a device that made long-term modifications of behavior on the basis of random events would have easily been beaten out by one that looked for systematic trends. Likewise, a device that made long-term modifications on the basis of experiences with parents or siblings would have been beaten out by one that used data with better predictive value: data collected in the arena in which the child will compete in adulthood.
Most animals, including humans, distinguish between relatives and nonrelatives; they favor relatives when it comes to sharing and nonrelatives when it comes to mating. If the relative–nonrelative distinction has already been made, why not use it to put a lower value on input coming from parents and siblings? This is only a suggestion; perhaps all that happens is that input from parents and siblings is outnumbered and outweighed by input coming from other sources. But there’s no theoretical reason why the system couldn’t distinguish between the various sources of input and rely more on some sources than on others.
The mental system I am proposing is tuned to receive certain kinds of signals from the environment: signals that help the individual work out a strategy of behavior that can be tried out in childhood and perhaps discarded if the results are disappointing or a more promising strategy becomes available later. This system requires a lot of brain power and takes a long time to complete its development, but human children are well provided with both.
Good-looking people receive better treatment in infancy, childhood, and adulthood. They are given more attention and get more respect from everyone—parents, teachers, peers, and employers—which means that there is widespread agreement on who is good-looking and who is not, and an equally widespread tendency to judge books by their covers.40 No, it’s not fair, but evolution cares not a whit for fairness.
As you might expect, good-looking people tend to be more assertive and sure of themselves. The female subjects in one experiment were treated rudely and made to wait while a researcher left the room in the middle of a fake interview. The less attractive women sat and waited for an average of nine minutes before raising a protest; the more attractive women started to complain after three minutes and twenty seconds. What produces the assertiveness is not the good looks themselves but the social effects of being good-looking. In childhood and adolescence, pretty girls have higher status among their peers; in fact, the effects for girls of being pretty are similar to those for boys of being tall and strong.41
The question is, how do children find out about their status, when it’s not simply a matter of who can dominate whom? In an earlier chapter I told you about attention structure: the high-status members of a group—children or adults, humans or apes—get looked at more often than those of lower status. I also told you about the eye-direction detector, a mental module that Simon Baron-Cohen included in his model of a mechanism for reading other people’s minds. The gaze detector, as I call it, gives a little beep whenever it notices that someone is looking at its owner.42
Being looked at by others makes a person more self-confident and more likely to speak out in a group. Researchers in a media lab in Canada were testing video conferencing equipment that enabled each participant in a group discussion to see images of the other participants. The researchers manipulated the images so that sometimes they appeared to be looking directly at a given participant—the subject—and sometimes they appeared to be looking elsewhere. The more gazes a subject received from these images, the more he or she contributed to the discussion. The timing of the gazes didn’t matter much; nor did their source. It was the sheer number of gazes the subject received.43 A simple counting device, working below the level of consciousness, determines whether you will sit through a meeting with your mouth shut or blurt out something you may later regret.
The counting device is one of the components of the status system; it counts eye-gazes the way a voting machine counts votes, and for much the same reason. This device is not new; vote-counting, like averaging, has been observed in nonhuman animals. Two British biologists recently presented evidence for what they called “group decision-making in animals.” Animal groups, they reasoned, are generally better off making decisions on a democratic basis, rather than following a leader, because the group as a whole is likely to have more and better information than any single animal within it. So decisions such as when to move and which way to go may be made by a majority-rules rule—in other words, by counting votes. African buffalo do this by tallying eye-gazes: the herd moves in the direction in which the majority of the adults are looking.44
Counting the eye-gazes they receive is one way that children and adults can assess their status; this is one of the cues the status system is sensitive to. But eye-gazes are not enough. Human social life is complex; status is multidimensional; the strategies available to children are limitless. They need information of a more subtle and detailed sort, information that cannot be conveyed by a simple vote-count.
What children need, ideally, is information on how they are seen by other people—what the “generalized other” thinks of them. They need a way of peering into other people’s minds and finding out what these people think of them.
We already know that normally developing children—children not afflicted with the disorder called autism—can do some mindreading. By the age of four, a child can figure out that if Anne hid the ball when Sally wasn’t looking, then Sally doesn’t know where the ball is. By around this age a child can also figure out that lying about what he did or didn’t do can put false information about his activities into another person’s mind, which means he knows that one of the things that other people can think about is him.45
Part of our ability to read other people’s minds—a very important part—is the ability to discern what other people think of us. It develops slowly, never becomes highly accurate, and some are better at it than others, but most adolescents and adults have a reasonably good idea of how others are reacting to them. We can usually tell whether someone we’re talking to likes us or not, respects us or not, and does or does not find us sexually attractive.
True, other people sometimes send out misleading signals. But the very fact that the signals are misleading shows that we must have a way of receiving and interpreting them. The signals come in a variety of forms: eye-gazes, facial expressions, body language, actions, and words. It is not unusual for mental organs to combine information coming in through different modalities; the mechanism that tells us where things are located in three-dimensional space can make use of vision, hearing, and/or touch.
Earlier I mentioned a set of mental devices called sociometers. The theorists who thought up these devices gave them a simple job: all they have to do is indicate, with a thumbs-up or thumbs-down signal, how a particular aspect of social life is going. The device I’m now proposing as a component of the status system is much fancier than a sociometer; it collects and processes infor
mation of a richer sort and issues more detailed reports. It can handle the thumbs-up/thumbs-down business—so it is, among other things, a sociometer—but it does harder jobs as well.
Here’s how I think it works. The status system, designed to collect and store information about the self, makes clever use of features of the relationship system (the system designed to collect information about other people). The activities of the two systems dovetail like this: while your relationship system is gathering information about me and storing it on the page assigned to me in your people-information lexicon, my status system is trying to figure out what you’ve got recorded on that page. You keep the information you’ve learned about me separate—you don’t mix it together with information about other people—but I take the information I’ve gotten about myself from my page in your lexicon and put it together with similar information I’ve gotten from other people’s lexicons. What I need is a picture of myself from the point of view of the “generalized other.”
Did that sound impossibly complicated? It will make more sense, I think, if I go through the series of steps by which evolution could have produced the mental system I’m proposing. It started back when animals became able to recognize and remember specific individuals of their species—their mates, their offspring, their adversaries in one-on-one dominance contests. Once that ability existed, another step became possible: now the animal could collect information on specific individuals and store this information in a mental lexicon. A useful ability for any social species.
The next innovation probably occurred fairly recently: during the six million years of evolution that separate our line of descent from the chimpanzee’s. What happened was that hominids gradually developed the ability to read each others’ minds. They got better and better at mindreading. This was undeniably a handy ability to have, but for a long time the information was used only by the relationship system.
Okay, now we have a mindreading mechanism plus a mental lexicon containing people-information. Put ’em together and what’ve you got? The logical next step: the ability to read what someone else has recorded about you on the page assigned to you in their mental lexicon. How handy is that?
I’m proposing a device that figures out what other people are thinking of you and how they regard you, and that puts together information collected from many different sources to give you a picture of yourself. Unfortunately the device doesn’t work perfectly—the picture is blurry—which is why the poet Robert Burns demanded his money back. “Oh wad some power the giftie gie us, to see oursels as ithers see us!” he memorably complained. The power is natural selection, and it did its best to gie us that giftie but was forced to make compromises. The reason we can’t read the page as accurately as we would like (or as accurately as we think we would like) is that the mind in which that page resides doesn’t want us to. It is to my advantage to know what you are thinking about me, but it may be to your advantage to keep me from knowing it.46
The picture may be blurry but it’s nuanced and multidimensional. The status system uses this information to work out a long-term strategy of behavior tailored specifically for its owner. Using data collected in childhood and adolescence—How many people can beat me up? How often do other people look at me? Do people trust me to give good advice?—the status system shapes and modifies personality in a way that takes account of the individual’s preexisting characteristics and the opportunities afforded by the environment. It places a bet—the stakes are high and sometimes it loses—that this is the best way to go, the best thing to do, to maximize the chances of success in the adult world.
The system I’m proposing wouldn’t work very well without the mindreading mechanism. It could rely on counting eye-gazes and the number of times its owner wins or loses dominance contests, but it’s hardly worth making long-term modifications of behavior on such a basis. Therefore I don’t expect animals that lack a mindreading mechanism to have a mental organ analogous to the human status system. I’m predicting, in other words, that nonhuman animals—even chimpanzees—lack a mental system that makes “personality” modifiable by experiences such as having high or low status in the juvenile group. In nonhuman primates, the behaviors associated with being an alpha male may be context dependent; being at the top for a while may have no lasting consequences. That may sound unlikely to you but it makes perfectly good sense for a nonhuman. After all, an animal’s status in its dominance hierarchy is subject to radical changes—gradual or abrupt—over the course of its lifetime, and the animal has to accustom itself, slowly or rapidly, to these changes. According to the primatologist Frans de Waal, when an alpha male chimpanzee is deposed he immediately loses his air of lordly self-confidence.47
Humans, too, have to accustom themselves to changes in status. So why did evolution provide our species with a system that shapes personality on the basis of childhood and adolescent experiences? The answer is that there came a point in hominid evolution when it became advantageous to be able to modify patterns of social behavior on a long-term basis. That point was reached when the mindreading mechanism was able to provide more subtle and multidimensional information than could be conveyed by eye-gazes or the outcomes of dominance contests, when individuals were able to make more subtle and multidimensional adjustments in their behavior, and when longer preparation was needed to fill specialized niches within the group.
I have described a mental system that shapes or modifies personality in a way that takes account of the individual’s size, strength, beauty, and so on. The existence of such a system has previously been suggested by others;48 what I’ve done is to show how it might work. The system I’m proposing is built out of components for which there is already good evidence and for which analogs exist in mental systems serving other functions, such as language.
Now for the big question. Can this mental system be the perpetrator I’m looking for? Can it be responsible for the personality differences between identical twins?
In the mystery Five Red Herrings, the detective Peter Wimsey eliminated five likely looking suspects. Then he was faced with a obstacle: the sixth suspect had what appeared to be an unassailable alibi. Before Wimsey could solve the mystery, he had to break down that alibi. He had to show that the sixth suspect could, in fact, have committed the crime.
There’s an obstacle in my case, too. The status system I’ve described is capable of shaping personality in a way that takes account of an individual’s size, strength, beauty, and so on. The problem is that identical twins are very much alike in size, strength, beauty, and so on. Therefore, other people will react to them in much the same way. The twins will have similar social experiences and will receive similar social feedback. Of course, they won’t have identical experiences, but their experiences should be more similar than those of two people who aren’t identical twins. And if their experiences are similar, the status system should make them more alike, rather than less alike!
But the problem, like the sixth suspect’s alibi, turns out not to be one. No, I’m not going to deny that the social experiences of identical twins are more similar than those of two people who aren’t twins—I admitted it a long time ago, when I was talking about the equally-similar-environments assumption in chapter 2. Nevertheless, the status system could still be the perpetrator we’ve been looking for. This is going to take some explaining, but I will be quicker about it than Peter Wimsey was.
To begin with, let’s look closely at the way behavioral geneticists apportion the variance in personality. Identical twins (who share 100 percent of their genes) are more similar in personality than fraternal twins (who share only 50 percent). Behavioral geneticists use this greater similarity as one way to estimate heritability—the amount of variance accounted for by differences and similarities in genes.
If personality is affected by social experiences, and if social experiences are in part the result of highly heritable characteristics such as size, strength, and beauty, then it’s true that the social experie
nces of twins are likely to make them more alike in personality. But this greater similarity is going to be included—wrongly, in the view of the critics of behavioral genetics—in the estimate of genetic effects on personality. That’s why estimates of the heritability of personality may be a little inflated; perhaps “true” heritability—the direct results of the actions of genes—is only .30 or .35, rather than .45. The behavioral genetic methods currently in use don’t permit researchers to distinguish between the direct effects of genes and the indirect effects, the effects of being treated in a certain way because of one’s inherited characteristics. It’s a serious flaw, according to the critics of behavioral genetics.
But, as I said in chapter 2, it’s no skin off my nose. On the contrary: it makes things easier for me. I’m interested in the personality differences between identical twins. The fact that they are treated more similarly by others cannot account for their differences: it can account only for their similarities, and their similarities are attributed to heritability. My interest is in the portion of the variance not attributed to heritability. The alleged flaw in behavioral genetic methods—the overestimation of heritability—actually simplifies my job, because the portion of the variance I’m interested in, the unexplained variance, has already had the effects of the more similar treatment of identical twins subtracted out of it.
No Two Alike Page 30