Here we encounter the primate-centric bias I mentioned in the introduction. What researchers who took the social-intelligence route were doing was looking at apes, seeing what their strongest suit was, and then assuming that human ancestors just took that suit a step further. They seemed not to see that the argument could just as easily be stood on its head. If apes were already so good at handling social relations, how could a handful of words or signs have improved their capacity? Given that human societies are more complex than ape societies, was it a case of “Societies got so complex that we had to have language in order to cope with those complexities,” as these researchers suggested? Or was it rather a case of “We got language and that was what made our societies more complex than theirs”? The second is at least as likely to be true as the first.
And in any case, we have to face the same problem we faced with tools and hunting. We’re being asked to believe that language, something radically opposed to existing means of communication, came along to help our ancestors do something they already did.
Variants on the “social intelligence” theme quickly diversified, and there’s at least one that’s coherent enough to merit discussion here.
This is Robin Dunbar’s “grooming and gossip” theory. Grooming, which of course includes delousing, is not just a hygienic but also a highly social activity among primates. It’s what bonds primates together, enabling primate societies to run (relatively!) smoothly. But it takes time. And if the social group gets too large, grooming, necessarily a one-on-one procedure, takes too long. You don’t have time to groom everyone you need to groom and still find enough to eat. So Dunbar suggests that language evolved as a grooming substitute; you can only physically groom one person at a time, but you can verbally groom three or four at a time. And, as Dunbar points out, a lot of our day-to-day chatter consists of a kind of verbal grooming—“buttering people up,” as they say.
Why couldn’t grooming have consisted of pleasant noises, without meaning—music, in other words? Because, to be efficient, verbal grooming had to hold interest, and what’s more interesting than gossip about other group members? Dunbar’s students went out to study social conversation and found indeed that most of it was personal gossip. Gossip as grooming, he reasoned, was therefore the original as well as the commonest modern function of language.
Dunbar’s theory sounds intriguing and initially plausible. It also escapes the need condition that trips up so many other proposals. If he’s right, and group size did increase, then protohumans were faced with a new problem that could well have required an equally novel solution. But did group size really increase among our ancestors? We don’t know (yet, anyway—someone somewhere should be working on it). We don’t even know, in fission-fusion societies such as chimps have and our ancestors probably had—fluid clusters that are constantly splitting and regrouping—quite what group size means or how it should be measured. And there are some other problems with the theory that we’ll touch on later.
For now, all we need to note is that it fails the ten-word test, what you might call the test of immediate utility. There just aren’t that many items of gossip you could pass on with ten words or fewer. And if you use most or all of your words for one particularly juicy item, say “Ug seduced your favorite female last night” (assuming, probably counter to fact, that a sentence with even this small degree of complexity could be produced and understood) what will you do for an encore? Keep repeating it? Novelty is the soul of gossip. But there’s no way in which a tiny number of words can be permuted to express a wide range of new items. You’d need at least several dozen, more likely a few hundred words before you could begin to do that. But you’d never get that far unless the first few words had already had a substantial payoff.
TESTS THAT THE THEORY MUST PASS
The test of immediate utility isn’t the only condition that an adequate theory of language origins has to meet. There are at least four others, and this seems as good a time as any to say what they are:
• Uniqueness
• Ecology
• Credibility
• Selfishness
Let’s look at each one in turn.
Uniqueness is there because any serious theory of how language began has to explain not only why language was acquired by humans, but why it wasn’t acquired by any other species. Even that isn’t quite enough. It should explain why, while language has developed to an extraordinary degree of complexity in humans, there isn’t a smidgen, not the least sign of it starting to develop, in any other species. Surely a unique effect requires a unique cause. But if the proposed trigger for language is anything that affects other species, it’s not likely to be the right one.
This immediately knocks out some otherwise promising suggestions.
For instance, it’s been suggested, by Geoffrey Miller among others, that the selective force driving language was female choice—a tried and trusted pressure, given the Darwinian seal of authenticity by the master himself, and confirmed in modern times by both observation and experiment. It explains, for example, why peacocks have such enormous, entirely dysfunctional tails. Females like them. They figure, “If he can survive with a tail that size, he must be hot stuff.” And, sure enough, if you trim a peacock’s tail he doesn’t score so often.
Does that work for humans? Well, there’s evidence on both sides. On the one hand, as songwriters Johnny Burke and Jimmy Van Heusen pointed out, “You don’t have to know the language” provided that the moon is shining and the girl has that certian expression in her eyes. On the other, there’s John Wilkes, eighteenth-century radical activist and notorious debauchee, whose face was ravaged by smallpox. “You’re so ugly,” a friend remarked, “how do you get all these women?” “Give me but half an hour with one,” Wilkes replied, “and I will talk away my face.”
But no matter who’s right (and as I wickedly pointed out in an article once, if female choice really worked with language, the captain of high school debating should score more chicks than the captain of the football team), the answer’s neither here nor there. The eloquence of a Wilkes and the mumblings of a protolinguistic protohuman are apples and oranges. Nobody doubts that, once language had really and truly gotten off the ground, it could, sometimes at least, enhance the evolutionary fitness of its most skilled users. The same is true for almost any language-origins proposal, the “language is power” one for instance: leaders often owe their positions to the gift of gab, and since, as Henry Kissinger told us, powerful equals sexy, they tend to get multiple mates too.
The trouble with all these explanations is, they involve things that a wide range of species share, certainly those nearest to us. In any number of species, females determine who they’ll mate with and pick those they think best. In many, probably most, primate species, individuals seek to enhance their status and scheme to obtain more power over other members of the group. If these factors operate in so many other species, why didn’t they lead to language in those species, too?
Moreover, none of these factors could work unless it first has something to work with. All of them, female choice, power-seeking, and the like, would surely have driven language once language had started. But how could they have started it? Female choice has to have something to choose from—in this case, presumably a whole range of different skills in using language. Seekers of power and status had to have something to seek them with; they would have to work at the high end of a range of language skills broad enough to have a high end. So none of these things could have had anything to do with the actual birth of language.
The second condition—ecology—merely means that explanations of language origins mustn’t conflict with what we know or can deduce about the ecology of our ancestors. That includes evidence drawn from the fossil and archaeological records, which are admittedly scanty and can seem contradictory at times. But this is no excuse for violating the ecology condition.
One thing that amazes me about the language evolution field is how o
ften people ignore this condition. Primate-centrists are the worst culprits. Since apes make such convenient and accessible models, and since we share so much of their DNA, primate-centrists assume proto-humans must have behaved pretty much as modern apes do. If there seem to be big differences now, well, recent civilization has just made us don disguises and draw veils over our basic ape nature.
As we’ll see in chapter 6, that’s a long way from the truth. Our remote ancestors may not have been much smarter than their ape cousins, but they lived in dramatically different environments and made their living in completely different ways. Unless you believe in universally uniform genes that impose identical behaviors wherever they occur—something modern biology has decisively refuted—you have to realize that partially arboreal, forest-dwelling apes make pretty poor models for the ways in which protohumans behaved.
As for the third condition, credibility . . .
It’s London, spring of 1998, the Second International Conference on the Evolution of Language. First thing to hit me was the cropped cannonball head and unapologetic London accent of the sociologist Chris Knight, asking without any preamble whatsoever:
“What does your theory have to say about the problem of cheap signals?”
“Er . . . duh,” I riposted eloquently.
I was coldcocked and blindsided. Cheap signals? What the hell were they? But Chris knows what he’s talking about, so I boned up fast, and this is the problem.
In the 1970s game theory began to be applied to biology. Wouldn’t it be the case that in a population where each individual strove for its own genetic success, there’d be payoffs for cheats and deceivers? Animals that exaggerated their prowess as potential mates might thereby gain access to breeding opportunities they’d never get through honesty. How could any animal know that the signals she was receiving really meant what they seemed to mean?
The Israeli biologist Amotz Zahavi came up with an answer. The harder a signal is to fake, the more likely it is genuine. Anyone can perform a fancy strut for a while, but to permanently carry around a peacock’s tail or the spread of a stag’s antlers suggests that their owner really is strong enough to sire sturdy children. In other words, signals had to be costly to be credible.
People like Chris were quick to pick up the implications of this for language. Words are notoriously easy to produce. Everyday speech is full of remarks to this effect: “Talk is cheap”; “He talks the talk, but does he walk the walk?”; “Sticks and stones may break my bones, but words will never hurt me.” Words are cheap tokens, so why should anyone believe them? This thought could not help but resonate at a time when, with Byrne and Whiten’s “Machiavellian strategies” center stage, everyone knew that primates in particular were adept at deceiving one another, even before words came along. But if no one would believe words, whence came the impetus to make up first tens, then hundreds, then thousands of the things?
Like the utility condition, the credibility condition strikes hardest at the very earliest stage of language, and suggests that, at this stage, language could never have gotten off the ground if the content of its first words couldn’t be verified immediately and beyond doubt. This, among other things, puts another nail in the coffin of the grooming and gossip theory—even today, we don’t believe half the gossip we hear.
Finally, selfishness. Over the last half of the last century, biologists shifted from believing that, at least sometimes, organisms did things for “the good of the species” or “the good of the group” to believing that everything every organism did was for itself, or at best for the genes it shared with close kin. The first point of view, known as “group selectionism,” quickly became a scientific no-no, inviting obloquy and scorn in equal proportions, though nowadays it’s slowly beginning to inch its way back in. (It’s fascinating to watch these cyclic to-and-fros in science, like the rise and fall of hemlines, but more stimulating, intellectually at least.)
But it’s too early yet to junk selfish genes. What look like behaviors “for the good of the species” may well turn out to be purely self-serving behaviors that happen, quite incidentally, to benefit the species as a whole. Dubious though it may be when pushed too far, the selfish gene notion has lit up too many areas of behavior to be lightly tossed aside.
So consider any linguistic act from this perspective. A gives information to B. Before the act, that information belonged to A exclusively. A could have exploited it for A’s own benefit. Now that’s not possible. B can exploit it too. What does A get out of this? If the answer is “nothing,” or even a negative payoff—A has to split a favorite tidbit with B—why should A transmit the information in the first place? If the answer is “B will repay the favor,” what guarantee does A have that B will reciprocate, won’t prove a cheat?
In other words, the first linguistic acts, whatever they were, must have been such that the speaker derived (at least!) as much benefit from them as the hearer did.
THE BIG-BRAIN FALLACY
Having reviewed the four tests—uniqueness, ecology, credibility, and selfishness—that any candidate theory of language evolution must pass, let’s dispose of the theory that as brains got bigger, our ancestors got cleverer and cleverer until they finally became clever enough to invent language.
This, in some form or other, is widely believed by many highly qualified scientists. For instance, in a recent interview, Nina Jablonski, who is not only head of anthropology at Penn State, but also, according to The New York Times, “a primatologist, an evolutionary biologist and a paleontologist,” explained that “in order to survive in the equatorial sun, [early humans] needed to cool their brains. Early humans evolved an increased number of sweat glands for that purpose, which in turn permitted their brain size to expand. As soon as we developed larger brains, our planning capacity increased, and this allowed people to disperse out of Africa.”
What’s wrong with this eminently plausible-sounding story? Lots of things. One, plenty of animals survive in the equatorial sun without increasing the number of their sweat glands. Two, permitting brain size to expand is not the same as obliging brain size to expand. Jablonski makes it sound as if brains were just bursting to expand but were held down by dumb stuff like inability to sweat properly. That’s far from true. Brains are highly expensive in terms of energy; animals have brains just big enough to do what they have to do, and anything over that is dysfunctional. Three, no one, to the best of my knowledge, has ever shown a correlation between brain size and planning capacity, for any species, least of all for species ancestral to humans, about whose planning capacities we know zero. Four, you don’t need any special planning capacity to get from one continent to another. All you need is a land bridge and feet—thousands of species have done it, most notably the placental predators of North America who, as soon as Central America was in place, exterminated the indigenous marsupial species of South America.
A dog has a bigger brain than a frog, and a dog can do lots more things than a frog can. This can only be because its bigger brain makes it smarter, you might suppose. However, a quarter century ago, the Scottish psychologist Euan Macphail wrote a paper that no one could challenge but everyone could and did ignore, which claimed that when you considered not the range of things animals could do but the actual mental apparatus with which they did them, there were only three levels of intelligence. There were organisms that could associate a stimulus with a response. There were organisms that could in addition associate a stimulus with another stimulus; all vertebrates and even some invertebrates fell into this class. And there were humans, who happened to have language. Macphail didn’t know how language made us more intelligent, but you will, if you read to the end of this book.
What is intelligence, anyway? Before we could compare the intelligence of one species with another, we’d have to have a valid definition and a valid measure that, unlike IQ, would work across species. Nobody has yet produced one. So even if there were more differences between the intelligence of species than Macphail c
laimed, there’s no way we could show by any nonsubjective means that one species was smarter or more stupid than another species.
If, as some have claimed, language was an invention of folk with big brains, it would be doubly unique. In addition to being the only system of its kind in nature, it would be the only biologically based behavior that had ever been consciously and deliberately created. And if you believe we can deliberately create biologically based behaviors, I have a couple of bridges you might contemplate purchasing.
But the real clincher is this. Brains don’t grow by themselves, of their own volition; they grow because animals need more brain cells and connections to more effectively carry out any new things they are beginning to do. In other words, brain-size increase doesn’t drive innovation—innovation drives brain-size increase, and in chapter 5 I’ll show you in detail how the process works, courtesy of the new and exciting theory called niche construction (it’s changed my whole way of looking at evolution, as I hope it will change yours).
It follows that we didn’t get a bigger and better brain that then gave us language; we got language and that gave us a bigger and better brain.
SO HOW COULD LANGUAGE HAVE EVOLVED?
Adam's Tongue: How Humans Made Language, How Language Made Humans Page 4