KINDS OF SIGNS
Surely it might. But what kind of signal would it have to be? So far, I’ve only mentioned two kinds: indexical and symbolic. But symbols can’t just pop out of the woodwork; unknown in any known ACS, they would have to be worked for and won in the earliest stage of protolanguage. And indices are irredeemably bound to the here and now, since they must point directly at whatever they refer to.
Fortunately, there is a third class, the iconic. An iconic sign is something that resembles what it refers to—in some way. It can be a part of the thing referred to, or a picture of it (or of part of it), or the noise it makes—anything that somehow evokes an object in the real world (or even an abstract class, as symbols do, it turns out).
I’m going to treat these three—icon, index, symbol—quite differently from the way Deacon treats them in The Symbolic Species. According to him, they form a hierarchy: icons at the bottom, indices in the middle, symbols at the top. Once or twice, when I’ve been thinking sloppily, I’ve endorsed this view. But viewed from the perspective of displacement, they’re not a hierarchy at all.
First of all, let’s make it quite clear that not all words are symbols.
Words can be iconic. There’s nothing in “dog” or “cat” that, in and of itself, evokes those particular animals. But words like “buzz” and “hiss” evoke by mimicry the sounds they describe. Moreover, “buzz” doesn’t necessarily refer to a particular noise on a particular occasion, in the way that iconic signs in ACSs do. A bristling of the fur is an iconic sign that can only mean a particular animal is angry at a particular place and time, but “buzz” may be applied indiscriminately, to refer to the noise made by the particular bee that’s bugging you right now or the generic noise that bees, wasps, beetles, and other insects habitually make—even the similar noise a crowd of people in animated conversation makes.
Words can be indexical. “This” and “that” are used exclusively to point to particular objects in the world. Unfortunately, they’re not very informative and can only be used if their referents have been established in the course of previous speech or writing, but they clearly can’t stand in for a specific class of real-world things, like “dogs” or “tables,” in the way a symbolic word can.
Then there are all the grammatical items that are also words: “the,” “a,” “who,” “not,” “by,” “for,” “of,” and so on. These, unlike icons, indices, and symbols, do not even refer; they merely establish relationships between words that do refer. So a simple statement like “Words are symbols” just doesn’t cut it.
It remains true, however, that most words are symbolic, and that without symbolic words we couldn’t have language. So here it comes, the question that The Symbolic Species never answered (or even asked, come to that)—where did symbolic words come from?
Well, think back to “buzz” and “hiss.” Compare these:
There’s a mosquito buzzing in my ear.
Nothing is more irritating than a buzzing sound.
In the first, there’s a particular buzzing in the here and now. In the second, there may be, but there needn’t be—I could say this in reacting to a story about something that happened years ago. In talking about symbolism and words, people often make far too much of arbitrariness—the absence of any relationship between a word’s form and its meaning. You almost feel there’s some kind of class distinction in there—up top are the symbolic words, words whose meaning you couldn’t possibly guess, and underneath the second-class citizens, words that shamelessly wear their meanings on their sleeves. But just as members of all social classes share the same bodily functions, so symbolic and iconic words share the same capacity for displacement. And when it comes to how language began, displacement is a factor far more important than arbitrariness.
Indeed, arbitrariness is a feature of many animal signals. To come back to the vervet’s leopard call, there’s nothing in it that, in and of itself, evokes or even suggests a leopard.
Rather than forming a hierarchy, symbols, icons, and indices can best be visualized as the points of a triangle. Indices, at one corner, cannot under any circumstances have a capacity for displacement. Symbols, at the remotest corner, must have that capacity. And icons, at a corner closer to symbols, may or may not have it depending on how they’re used. Among other animals, icons never developed that capacity, but it was potentially there, and it blossomed when language came along.
Iconicity, therefore, is the most probable road that our ancestors took into language.
So we’ve now completed the task that, as language engineers, we set out to accomplish. We can specify precisely the conditions under which an ACS could begin to morph into something completely different, with different goals and different means for getting there. Here are the specifications:
Selective pressure: the need to transmit information about food sources outside the sensory range of message recipients.
Probable means: iconic signs.
Now the question becomes, can we find any evolutionary model for this? Are we talking about something wholly unique to some human ancestor, or is there anything comparable among other species that we could use to guide us here? If not, we run the risk, endemic to this field, of floundering in a morass of speculation, a morass with no empirical floor to prevent us from being swallowed by the mud. It’s happened to all too many before us.
So let’s stop thinking like engineers and start thinking like biologists. The comparative method is the core of evolutionary biology, and it’s precisely the absence of anything language could be compared to that has bedeviled the field from the outset. It’s time to start looking at evolution and how it has worked in the past, to try to make some linkage between our specifications and things that have really happened.
Where to start, though?
The obvious place, the place almost everyone starts, is among the great apes. After all, they’re our nearest relatives, made from almost the same genetic material. If there’s continuity in the development of language, that continuity surely should have a straightforward genetic basis of some kind.
Well, in the next couple of chapters we’ll see about that.
3
SINGING APES?
THE IMPORTANCE OF BEING PRIMATES
By the close of the nineteenth century, everyone who wasn’t blinkered by religious dogma knew that humans, far from being the special-purpose product of some interfering deity, were members of the primate family who could only have evolved from something resembling a chimpanzee. The full significance of this finding wasn’t appreciated until in 1954 the discovery of the double helix structure of DNA ushered in the century of the gene.
As discovery followed discovery, genetics came to dominate the life sciences. The determinism of Richard Dawkins’s The Selfish Gene went mainstream; while lip service was paid to the environment and its influence, a consensus grew that genetics formed by far the most important driving force in evolution. As the close resemblances between the DNA of humans and great apes became apparent, more and more people assumed that most if not all of what had been seen as typically human (and in many cases, uniquely human) traits and behaviors were no more than expansions of traits and behaviors found among the apes. Such a view was encapsulated in the titles of popular works on human evolution: The Naked Ape, The Third Chimpanzee, and (of course) The Ape That Spoke and The Talking Ape.
Here lie the roots of what Irene Pepperberg described as the “primate-centric bias” in language evolution studies. It seemed self-evident that if humans were, indeed, no more than souped-up apes, the origins of language, or at least its immediate precursors, had to be found among chimpanzees, gorillas, and orangutans. This belief was reinforced by attempts, from the 1960s on, to teach some form of language to these species (more on this in the next chapter). Although, to most linguists, results of these experiments were equivocal at best, the case made by “ape-language” researchers found many supporters in other behavioral sciences, and has been strengthe
ned, over the past couple of decades, by the performance of another great ape, the bonobo or pygmy chimpanzee.
Certainly the simplest and most straightforward story of how language evolved would go something like this:
Five to seven million years ago, the primate line of descent (which had already branched to throw off first orangutans and then gorillas) divided again. One branch went on to father bonobos and chimpanzees; the other was that of our ancestors. These ancestors, driven out of the forest and into the grasslands by global drying, were forced to make meat a substantial part of their diet. A rich meat diet enlarged their brains and thereby increased their intelligence. Intelligence was also enhanced by a factor common to chimpanzees and bonobos—social competitiveness. Apes were always trying to outwit one another, to gain prestige in the group, preferential access to mates, first dibs in a monkey hunt. This factor selected for the intelligence that meat supplied, setting up a beneficent spiral. The old ape ACS, drawn into this spiral, expanded and diversified. At some point that it might be quite impossible for us to determine, the ACS flowed seamlessly into language. Language made life more complicated, and in turn itself grew more complicated to deal with these complications, until we finally arrived at our present situation.
To doubt that our language has its foundations in ape behavior has become, in some circles, almost a heresy, enough to brand the doubter a closet creationist. That this conventional wisdom has achieved the status of dogma, to be accepted on faith, regardless of the absence of evidence—even if what evidence there is points in a contrary direction—is made clear in a quote from Steven Mithen’s recent book on the evolution of language: “And yet it is in this African ape repertoire, and not in those of monkeys or gibbons, that the roots of human language and music must be found. The apparently greater complexity of monkey calls must be an illusion, one that simply reflects our limited understanding of ape calls” (my italics). Note that, as well as demonstrating an abiding faith in the conventional story, this passage inescapably suggests a bias I mentioned in the last chapter: homocentricity.
I suspect most of those who subscribe to the conventional wisdom don’t realize they’re homocentric; most likely they’d indignantly deny it, pointing out that, instead of placing humans center stage, they’re emphasizing all the things we have in common with apes. Maybe, but if you see apes as just slightly less complex versions of ourselves, you’re implicitly saying they’re merely a crude foreshadowing of humans—In the Shadow of Man, as Jane Goodall’s book title so aptly put it. They’re worthy of our attention only insofar as we can find in them forerunners of typically human traits—they’re “almost human.” And where language is concerned, it’s as if you had something like the old scala naturae of pre-Darwinian nature study, a ladder leading up through progressively richer and more complex systems of communication until finally, among our nearest relatives, you found something that needed only a tweak or two to blossom into language, or at least protolanguage. It’s an embarrassment to this whole approach that monkeys turn out to have things more wordlike than anything found among chimps and bonobos.
If ACSs formed a ladder to language, the species closest to us should have the most languagelike ACSs. But they don’t, and the reason they don’t is because ACSs are not failed attempts at language. They’re not crude and misguided attempts to do what we do. They are autonomous systems that exist to serve the adaptive needs of each species that has one. Species get what they need and no more, and in the section after next we’ll see good reasons why the ACSs of great apes should be, in some ways at least, less, rather than more, complex than those of monkeys.
SOME PROBLEMS WITH BEING PRIMATES
Let’s look at some of the things that are wrong with the conventional story.
Many versions of this story have language arising through an intensification of social competitiveness—a kind of arms race between increasingly smarter and more devious protohumans. But this ignores two vital things. It ignores the ecology of our post-chimp, prelanguage ancestors, plus all the changes from typically ape social behavior that this very different ecology would inevitably impose. And it ignores the fact that those ancestors developed, at some stage, but likely a long time ago, a type and degree of cooperation unknown in any other primate species. It’s not that humans aren’t socially competitive—they are, and highly so. But, paradoxically, they are also highly cooperative, capable of combining their forces for joint ventures in units that range from dyads to many millions of individuals. Among apes, on the other hand, the dyad is about the largest unit in which cooperation can occur among unrelated individuals; even then, it’s only of the “I’ll scratch your back if you scratch mine” kind. Both ecology and the source of cooperation will be discussed in chapters 6 and 8.
Then there is the dubious nature of the connection between language and intelligence, an issue that will come up in several chapters, but particularly in chapters 4 and 10. At first sight, that brain link looks like a no-brainer. We’re smart, we have language; other animals are less smart, and they don’t have language. We assume that intelligence is the broader category. Since we usually regard language as no more than the means by which we express our thoughts, it seems natural to think that language should issue from intelligence, rather than vice versa. It seemed equally obvious, to naive observers, that the earth was the center of the universe, and the sun, moon, and planets all went around it.
When it comes to mind, intelligence, and language, we’re just about where people were with regard to the universe, say a thousand years ago.
However, I’d like now to focus on the transition issue (that is, the transition from an alingual state to anything that you might, with a stretch, classify as the beginning of language) and examine some of the problems this transition must face if we adopt a purely primate-centric approach.
Most accounts gloss over this transition, one way or another. Where we need a precise, detailed, carefully reasoned analysis, we seldom get more than hopeful hand-waving. But the transition from no language to some kind of protolanguage is where the rubber meets the road—it’s the crux, the core moment in language evolution, and the whole of the middle portion of this book will be devoted to what I believe, in our present state of knowledge, is the best and perhaps the only possible explanation of that moment. But first we need to see why a straight-line evolution from a typical ape ACS to language cannot be maintained.
We should begin by seeing just what we have to work with. Accordingly, we’ll look at the current ACSs of bonobos and chimpanzees, assuming that these systems have not deteriorated or shrunk since their line split from that of humans. (We cannot, of course, know this for certain, but there’s no reason to suppose otherwise.) Since, in the first two chapters, I’ve described what the most minimal language should be able to do, we can then assess how plausible it is that some expansion or complication of these ACSs might lead to anything you could call protolanguage.
THE RAW MATERIAL OF LANGUAGE?
One of the best recent accounts of chimpanzee and bonobo ACSs is by Amy Pollick and Frans de Waal of the Yerkes National Primate Research Center. They list a total of thirty-one gestures, fifteen vocalizations, and three facial expressions. Of these signals, three gestures and six vocalizations were produced only by chimpanzees, while two gestures and six vocalizations were produced only by bonobos. From these data Pollick and de Waal draw a rather surprising conclusion: support for the theory that human language was originally gestural rather than vocal.
This is surprising because if 84 percent of the gestures are shared by the two species, but only 20 percent of the vocalizations, the gestures are more likely to be homologies—traits directly inherited from the common ancestor of chimps and bonobos, rather than innovations appearing after the split. On the other hand, vocalizations, since they differ more in the two species, must have developed mostly since the two species split. And, since language too is vocal and an innovation, we would expect protohumans, just like apes, t
o have been moving away from their last common ancestor, not toward it—shifting from gesture toward a more vocal medium.
But the claim that language started in gesture is not the only dubious claim Pollick and de Waal make. They note that gestures are more contextually flexible than vocalizations—that is to say, the meanings of gestures vary according to the contexts in which they are used. For instance, the gesture described as “gentle touch” may be an invitation to sex if used by a male in the presence of a receptive female, but a request for milk if used by an infant to its mother. On the other hand, the “scream” vocalization is always and only given (by members of both species) if an individual is under threat or actual attack. They therefore argue that gestures, like words but unlike calls, are not tied to particular situations, and thus show movement away from the rigid restrictions of ACSs.
But at least the calls show consistency of meaning, whereas gestures have no consistent core of meaning—unlike words, they take their meanings entirely from the circumstances in which they are used. (Imagine, if you can, a word “moosh,” which means “Let’s have sex” if uttered by a male to a female and “Give me milk” if uttered by a baby to her mother.) Gestures differ from calls only in that they are linked to several different kinds of situations rather than just one, and mean something quite different in each situation. That’s not how words work.
Moreover, the goals of all these gestures and vocalizations are anything but languagelike. They have two goals only. One goal is to express emotion. The “food peep” vocalization, for instance, is seldom if ever used to indicate the location of food to other group members—its “meaning” is “Yum yum!” rather than “Come and get it!” The other goal, as with the “gentle touch” discussed above, is to manipulate another member of the group. Monkey alarm calls at least give concrete information (or can be so interpreted), but not one call or gesture in the bonobo or chimpanzee repertoires transmits any objective information about the environment.
Adam's Tongue: How Humans Made Language, How Language Made Humans Page 7