The Mating Mind: How Sexual Choice Shaped the Evolution of Human Nature (53 page)

Language Through Kinship and Reciprocity?

Shared information is multiplied, whereas shared food is divided. By giving you a useful fact, I do not automatically lose the benefits of knowing it. Potentially, this information-sharing effect could have made it rather easy for language to evolve through kin selection and reciprocal altruism. Our ancestors lived in small, semi-stable groups full of relatives and friends. By evolving the ability to share information with them, our genes and our social relationships would have benefited.

This sounds useful, and it is probably mostly right. However, there are still conflicts of interest. Relatives do not share all of the same genes, so do not have identical evolutionary interests. Likewise for friends in a reciprocity situation: there is always the temptation to cheat by receiving more than one gives. Given these conflicts of interest, we can look at the costs and benefits of language to see whether people's real behavior follows the predictions of kinship and reciprocity models.

As long as language is viewed purely in terms of information transmission, it will be seen as bringing more benefits to the listener than to the speaker. The speaker already knows the information being conveyed, and learns nothing new by sharing it, but the listener does gain information by listening. Information is still like food in this sense: it is better to receive than to give. In the bare-bones kinship and reciprocity theories, the principal benefit of language must be to the listener. This leads to an interesting prediction: we should be a species of extremely good listeners and very reluctant talkers. We should view silent, attentive listening as a selfish indulgence, and non-stop talking as a saintly act of altruism. People should pay huge amounts of money to engage in the vice of being psychotherapists, who get to hear people's innermost secrets while having to reveal little of themselves.

This does not describe the human species as I know it. Watch any group of people conversing, and you will see the exact opposite of the behavior predicted by the kinship and reciprocity theories of language. People compete to say things. They strive to be heard. When they appear to be listening, they are often mentally rehearsing their next contribution to the discourse rather than absorbing what was just said by others. Those who fail to yield the floor to their colleagues are considered selfish, not altruistic. Turn-taking rules have emerged to regulate not who gets to listen, but who gets to talk. Scientists compete for the chance to give talks at conferences, not for the chance to listen. For psychotherapists to use the "non-directive" methods advocated by Carl Rogers—in which the therapist says nothing back to the client except paraphrases of what they have heard—requires an almost superhuman inhibition of our will to talk.

Nor do the kinship and reciprocity theories predict our anatomy very accurately. If talking were the cost and listening were the benefit of language, then our speaking apparatus, which bears the cost of our information-altruism, should have remained rudimentary and conservative, capable only of grudging whispers and inarticulate mumbling. Our ears, which enjoy the benefits of

information-acquisition, should have evolved into enormous ear-trumpets that can be swivelled in any direction to soak up all the valuable intelligence reluctantly offered by our peers. Again, this is the opposite of what we observe. Our hearing apparatus remains evolutionarily conservative, very similar to that of other apes, while our speaking apparatus has been dramatically re-engineered. The burden of adaptation has fallen on speaking rather than listening. Like our conversational behavior, this anatomical evidence suggests that speaking somehow brought greater hidden evolutionary benefits than listening.

Verbal Courtship

Much of human courtship is verbal courtship: "boy meets girl" usually means boy and girl talk. At every stage of courtship, language is displayed, and language is subject to mate choice. Teenagers agonize over the words they will use when they telephone someone to ask for a date. Stuttering, sudden changes in voice pitch, awkward grammar, poor word choice, and uninteresting content are usually considered such fatal errors by their perpetrators that they often hang up in shame, assuming that they will remain sexual failures forever. Things are not so different a little later in life. Adults in singles bars nervously rehearse their pickup lines, and mentally outline their conversational gambits.

After basic greetings, verbal courtship intensifies, progressing through self-introduction, observations concerning immediate social surroundings, compliments, and offers of minor favors. If mutual interest is displayed, people go on to trade more personal information, searching for mutual acquaintances, shared interests, and ideological common ground. If there is no common language or if accents are mutually unintelligible, courtship usually breaks down. At each stage, either person may break off courtship or attempt to escalate intimacy, but usually at least several hours of conversation precede even minor physical contact, and at least several separate conversations over several encounters precede real sex. This verbal courtship is the heart of human sexual selection. Although people may be physically

attracted before a word is spoken, even the most ardent suitors will offer at least a few minutes of verbal intercourse before seeking physical intercourse.

All of this is quite obvious to any adult human with a modicum of social experience. But whereas toddlers can learn to speak reasonably well within three years of birth, it usually takes at least a decade of practice before young adults are comfortable with the basics of verbal courtship. To an evolutionist interested in sexual selection, adolescence is fascinating. The 19th-century biologist Ernst Haeckel's claim that "ontogeny recapitulates phylogeny" is often misleading, but there are cases, especially in sexual selection, where stages of life-cycle development may reflect past stages of evolutionary history. The awkward, uneven, sometimes witty verbal courtship of teenagers may not be such a bad model for the verbal courtship of our ancestors during the evolution of language. There must have been some similarities: poor vocal control, small vocabulary, uncertainty about conversational conventions, difficulty in finding phrases to express thoughts. As every parent of a teenage boy knows, the sudden transition from early-adolescent minimalist grunting to late-adolescent verbal fluency seems to coincide with the self-confidence necessary for dating girls. The boy's same-sex friends seem to demand little more than quiet, cryptic, grammatically degenerate mumbling, even when playing complex computer games or arguing the relative merits of various actresses and models. Girls seem to demand much more volume, expressiveness, complexity, fluency, and creativity. If natural selection had shaped human language for the efficient, cooperative communication of useful information, we would all speak this sort of "Early Adolescent Mumbled Dialect." At least in males, only with the demands of verbal courtship do we witness the development of recognizably human-level language.

Computer pioneer Alan Turing alluded to the importance of verbal courtship for testing someone's mental capacities in the original 1950 version of his "imitation game," which has come to be known as the "Turing test." In the imitation game, an interrogator tries to determine whether he is interacting with a

real woman or a computer program that imitates a woman. Turing was more interested in intelligence than female flesh, so he eliminated the physical cues of womanhood, and limited the interrogator to typing questions on a terminal, and receiving answers on a screen. The questions can be as challenging as the interrogator likes, such as "Please write me a sonnet on the subject of the Forth Bridge." In Turing's view, if a computer can successfully lead an interrogator to believe that he is interacting with a. real woman, it should be considered intelligent. Turing emphasized that the computer must be capable of credibly demonstrating a very wide range of behaviors—his list included being kind, using words properly, having a sense of humor, catching us by surprise, claiming to enjoy strawberries and cream, falling in love, and making someone fall in love with it. (Strikingly, many of these behaviors overlap with the courtship adaptations we have considered in previous chapters.)

After Turing, philosophers of artificial intelligence dismissed the sexual aspect of the imitation game as a confusing distraction, and stripped it away from modern versions of the Turing test. However, Turing's original version subtly pointed to the special challenges of demonstrating human intelligence during courtship. Even a very simple 1970s computer program like ELIZA can fool people into thinking that they are interacting with a real psychotherapist—but no one has fallen in love with ELIZA, as far as I know. Turing's more sexualized imitation game offered a key insight: human intelligence can be demonstrated very effectively through verbal courtship, and any machine capable of effective verbal courtship should be considered genuinely intelligent.

The idea that language evolved for verbal courtship solves the altruism problem by identifying a sexual payoff for speaking well. Once the rudiments of language started to evolve, for whatever reason, our sexually motivated ancestors would probably have used their heritable language abilities in courtship. Language complexity could have evolved through a combination of runaway sexual selection, mental biases in favor of well-articulated thoughts, and fitness indicator effects.

Language Displays and Social Status

This verbal courtship theory fits nicely with some ideas developed by three other language evolution theorists—Robbins Burling, John Locke, and Jean-Louis Dessalles. They are not as well known as Noam Chomsky or Steven Pinker, but they share my belief that a good theory of how language evolved must show how selfish genes can derive hidden benefits from the apparently altruistic act of speaking. In an important paper published in 1986, anthropologist Robbins Burling advanced arguments similar to mine. He contrasted the excessiveness of our baroque syntax and enormous vocabulary with the sufficiency of simple pidgin languages for trade, hunting, and tool making, and considered this alongside the problem of language's apparent altruism. He proposed that complex human language evolved through male orators competing for social status by speaking eloquently, since high status would give them reproductive advantages. Burling cited anthropological evidence of the links in tribal societies between verbal skill, social status, and reproductive success. As long as those links held true during human evolution, language could have evolved ever greater complexity. As Burling noted, "All that is needed for the mechanism I suggest to be effective is that the average leader in the average society have slightly more verbal facility and slightly more children than other men." Although he emphasized verbal leadership more than verbal courtship, he did acknowledge that "We need our very best language for winning a lover." I think Burling's sexual selection model of language evolution deserves much more attention than it has received, and complements my ideas about verbal courtship.

Cambridge linguist John Locke has extended Burling's social-status model with more linguistic evidence, paying more attention to the role of "verbal plumage" in human sexual mate choice. He quoted from a study in which a young African-American man from Los Angeles patiently explained the sexual-competitive functions of language to a visiting linguist: "Yo' rap is your thing . . . like your personality Like you kin style on some dude by rappin' better 'n he do. Show 'im up. Outdo him conversation-

wise. Or you can rap to a young lady, you tryin' to impress her, catch her action—you know—get wid her sex-wise." In a few concise phrases, this teenager alluded to both classic processes of sexual selection: male competition for status, and female choice for male displays.

Along similar lines, language researcher Jean-Louis Dessalles has pointed out that listeners award higher social status to speakers who make relevant, interesting points in conversation. Language may have evolved through social selection to permit these "relevance" displays. This is why people compete to offer good ideas and insights when talking in groups. While Burling and Locke focused on dramatic public displays of oratorical prowess, Dessalles focused on social competition to say interesting things in ordinary small-group conversation.

Burling, Locke, and Dessalles have all identified important selection pressures that have been neglected in previous theorizing about language evolution. They have shown how language's hidden status and sexual benefits could have driven its evolution. In their theories, sexual attractiveness depends on social status, which in turn depends on verbal ability displayed in large or small groups. In my verbal courtship theory, sexual choice favored verbal ability more directly through one-to-one conversation. Sexual selection probably shaped human language in both ways: directly, through mate choice, and indirectly, through social status. Here I focus on verbal courtship only because it has received less attention so far.

A Million Words of Courtship

Verbal courtship can be quantified. Conception of a baby is the evolutionarily relevant threshold for success in courtship. Without contraception, it takes an average couple about three months of regular sex before a pregnancy occurs. If we assume two hours of talk per day in the early stages of sexual relationship, and three words spoken per second (an average rate), each member of a couple would have uttered about a million words before they conceived any offspring. Each would have talked enough to fill six

books the length of this one. In modern societies, the surprising thing is not that couples run out of things to say to each other, but that they do not run out much sooner.