Webbed Reprint Collection

---

William H. Calvin

University of Washington
Seattle WA 98195-1800 USA

9

Protolanguage Emerging

Why so long ago? Well, although our ancestors Homo habilis and Homo erectus were probably closer to apes in their behavior than they are to modern people, they still showed substantial differences from any other primate. They had brains bigger than those of apes, indeed brains that grew steadily throughout the last several million years to a size within the range of modern humans; they spread through a major part of the Old World; they made tools which, though crude and clumsy beside even the tools of Cro-Magnons, were far ahead of anything other species could manufacture; they knew and used fire; they built crude shelters against the weather; and they probably had other skills. Nature preserves only stones and bones, or things that can turn to stone. Artifacts of wood or fiber almost always decay and are lost to us (although recently, in a European bog, long wooden spears with sharpened points that have been dated to half a million years ago were found). None of these things makes it certain that they had protolanguage then but, given that their brains were not that much smaller than ours, it is reasonable to suppose that they did. And don=t forget, it=s really very unlikely that human language as we know it emerged recently and in a single piece. There must surely have been at least one stage intermediate between no language and full language.

That=s the question everyone was asking in the decade or two following the publication of Darwin=s On the Origin of Species in 1859. What they immediately focused on was the Magic Moment when language began, and they asked such questions as what were the first words like, what did they mean, for what purpose were they created? So much wild and baseless speculation briefly flourished that the linguistic community got turned off on the whole topic of language evolution for the century that followed; even today, most linguists show little interest in it. The focus of those early speculations was unfortunate, for there were many equally interesting and probably more easily answerable questions that could have been asked, such as, given that language got started somehow, how could it have progressed to the complex systems that all human languages possess today?

Probably people then didn=t think that question needed answering. Until well into the twentieth century, there was little or no serious study of syntax, certainly nothing beyond long, detailed lists of all the types of clause, phrase, and sentence structure that particular languages used. The big problem seemed to be getting people to talk at all. Once that had been accomplished, it was assumed that language would just naturally get more and more complex.

Nowadays, as subsequent chapters tell, the story of how language got from its birth to its present state is longer, more complex, maybe even more interesting than the story of how language got started. But that earlier story still needs to be told, even if we can=t be as sure of the details as we can for the later stages.

---

WHC: Any theory of cooperation suggests that, as a prerequisite, you have to be able to identify individuals. I=d suggest that unspoken proper names B and particularly for the individuals that you don=t see every day B might be a good start for evolving words. Every primate living in a social situation has the problem of identifying the others as individuals, simply because of dominance: does that individual chase me if I don=t give way? Or can I threaten him instead? You need to keep track of the state of mutual support among your fellows as well, to judge a social situation B and, as group size increases, the dyads to be remembered rise spectacularly (N!). So proper names are likely to have the conceptual structure there, ready and waiting in the tip of the temporal lobe, for whenever there was the need to make them into words (recombineable units) that were spoken or communicated by stares.

DB: Right, it could be that the naming of individuals was the very first use of words. It=s worth noting here the so-called Asignature whistles@ of dolphins: each dolphin in a pod has a whistle different from that of other pod members, and it seems these whistles are used like names, for identification purposes. But we simply don=t know. There=s a confusion here that=s often perpetuated by people using the term Anaming@ in an illegitimate sense. They talk about Anaming the things in the world@ or treat Adog@ as a name for Adogs.@ Most words are simply not names. A name is associated with a particular individual, a word identifies the concept of a whole class of people or things. So getting names doesn=t necessarily buy you words. Moreover, apes and monkeys already recognize one another as individuals and indeed have clear maps in their heads that tell them which individuals are related to one another, without any kind of language to help them.

---

Over the last couple of decades, at least four developments in the behavioral sciences have combined to focus attention on primate intelligence: ethological studies of primates, the concept of AMachiavellian intelligence,@ theories of Atheory of mind,@ and experimental projects with apes. It=s worth a quick look at all of these to see why they made social intelligence seem so attractive as the major selective pressure toward language.

It=s hard to remember that forty years ago, all that we knew about primate social life was derived from looking at primates in zoos. This was like studying people in jail and then generalizing to the whole human species. Fortunately, people like Jane Goodall and George Schaller initiated the study of primates in their natural environment, and the result of their pioneering work has been a spate of studies (not to mention shows on the Discovery channel) that have made us almost as familiar with the natural social life of apes as we are with our own. Indeed, the most remarkable thing about ape society is that it does so closely resemble our own. We see the same maneuvering for status, the same family feuds, we see parental and filial affection, the forming and reforming of alliances, altruism, cheating, loyalty, revenge, betrayal. About the only difference between them and us is that we tend to hide our emotions more, bear grudges longer, and accompany all our actions with a torrent of words. Language on top of primate social life is like a mural on a wall B it adds interest and decoration, but it really doesn=t change the structure of the wall.

The second development, which arose in part out of the first, was triggered by a seminal work by Nicholas Humphrey. Humphrey asked what selective pressure had served to raise the intelligence of primates above that of other families. The most plausible candidate was social life itself. In the small, tight-knit groups that characterized many (but not all) primate species, there would be a ceaselessly escalating arms race between individuals, each seeking to get the better of the others. Humphrey=s work was followed by a series of studies, in the main anecdotal but impressive in their number and consistency, showing acts of calculation and deception on the part of several primate species, which seemed to become more frequent among species closer to us. It seemed as though the tough act of staying successfully selfish within a social community, and competing with individuals some of whom were as smart or smarter than oneself, required a lot more intelligence than, say, hunting or making tools. So, if language was the ultimate flower of intelligence, what more plausible conclusion than that social intelligence produced language?

The third development was a growing interest in what is known as Atheory of mind.@ An age-old philosophical problem is, how do we really know what goes on in other people=s minds? How can we tell whether the concepts or emotions or subjective experiences they have are the same as, or different from, those that we have, and to which we all give the same names? One thing we feel sure of is that the knowledge in other people=s minds isn=t always the same as ours. But children aren=t necessarily aware of this. One test for Afalse belief@ goes as follows: the child is shown two puppets, call them Bib and Bob. Bib puts some food in a box, then goes away. While Bib is away, Bob moves the food to a second box. Bib comes back. The child is asked, AWhere does Bib think the food is?@ Most children under four will answer, AIn the second box.@ They know it, and Bob knows it, so why shouldn=t Bib know it?

Over two decades ago, the question was raised as to whether other species, in particular other primates, had a theory of mind. Today there is still no clear answer. However, a number of scholars believe that, like language, theory of mind is not an indivisible monolith. One component of a theory of mind is having some idea about what other individuals want or intend. But, as these scholars point out, this is a prerequisite for language as we know it. If someone says, AIsn=t it scary in here?@ we may know what the actual words mean, but we also need to know whether the speaker is appealing for our reassurance or simply wants us to confess our own fear. And because manipulating others (into, say, getting them to reassure you) or tricking them (into, say, getting them to confess weakness to you) are, at a simpler, nonlinguistic level, things that other primates already do, what more likely than that these competing drives (to trick and manipulate, versus avoiding being tricked and manipulated) constituted the selective pressure that favored language?

Finally, attempts to teach symbolic systems to apes, once their focus had changed from Adetermining whether apes can acquire language@ to Awhat components of language can apes require?@ provided a powerful reinforcement for the social intelligence theorists. If apes already had components of, or at worst, prerequisites for language, what likelier hypothesis than that the demands of competition within small self-contained societies brought these latent characteristics to the surface?

Yet in spite of this, there are good reasons for supposing that the first emergence of protolanguage had very little to do with social intelligence. I=ll start with some doubts about a very widespread assumption. One reason for theorists discarding hunting and toolmaking as prime movers in language is that these activities require relatively little intelligence as compared to the demands of a complex social life. Concealed in this claim is the assumption that to start any kind of language requires a lot of intelligence. Once exposed, the assumption looks far from certain. If creatures like sea lions and parrots can be taught symbolic units, then what a creature needs to acquire them spontaneously is surely not so much intelligence as a need to impart concrete information about the real world.

Because that=s what language does to perfection, while other animal communication systems can hardly do it at all. Animals can give alarm calls warning of a predator, but they can=t say, AThose footprints are a leopard=s@ B they can=t even point to the footprints and say Aleopard!@ Animals can erect their fur or show other signs of anger but they can=t say AI=m angry with you because you cheated me.@ We may use language most often to flatter people or exchange gossip, but that doesn=t tell us anything about why language arose in the first place. By the same reasoning, we could prove that computers were invented to surf the web or play video games. It makes no sense to try to explain language while ignoring what language does uniquely and does best.

AMore complex social lives@ would be a good answer, if the facts supported it. They don=t. As noted above, if you take language away, it=s far from clear that our social lives are any more complex than those of chimpanzees or bonobos. How would language have increased the fitness of animals who could already handle a rich and varied social life? If there had been no other motive, language might even have been dysfunctional. The animal most skillful in using it might have given away too much about itself, might have provided a more taciturn rival with information that could then have been used against it.

An increase in group size is another possibility that has been suggested. But there is absolutely no evidence that group size among hominids one million or two million years ago was any greater than group size among bonobos or chimpanzees. On top of which, it=s not clear that group size is a criterion of anything. Human hunter-gatherers and chimpanzees both have what are called Afusion-fission@ societies B their group size fluctuates as they divide into small groups and then periodically come together again in larger groups. Which group size is the right group size?

So the selective pressure for language had to come from something that was unique to hominids and something that required the exchange of factual information. Let=s think for a moment about hominid ecology and how it differed from that of apes, ancient or modern. Apes live mostly or entirely in heavily forested regions of the winterless tropics. The frequency of trees and the ability of apes to climb them better than any potential predator put apes virtually beyond the reach of effective predation. What this means in terms of their daily life is that they don=t have to devote the time and energy that many creatures do to the business of watching for and evading predators.

Chimpanzees aren=t strict vegetarians by any means, but in the tropical regions from which they have never succeeded in escaping, they can usually find enough fruit, nuts, or leafage to satisfy them. Meat is a rare luxury. They can usually locate an adequate supply of food without going very far. In the course of a day they may not wander outside of an area of a square mile or so.

On both these counts B predation risk and food availability B the lives of early hominids were very different. Instead of tropical forests, they inhabited grassy savannas with isolated stands of trees and what are known as Agallery woods@ B those narrow, winding woodlands, only a few trees wide, found on the banks of rivers in savanna areas. There one couldn=t be sure of a tree to run up, or of success if one did, as there=s an obvious tradeoff between bipedal walking and the ability to climb trees. It would do a modern human little good to try climbing a tree if pursued by a leopard, for instance. We don=t know how quickly or how gradually that tradeoff occurred, but we do know that our ancestors started walking on two legs at least a million years before any serious brain enlargement, so they were probably relatively poor tree climbers by the time even protolanguage emerged.

Moreover, savannas were, as they are today, prime predator country. Outside comic books, the lion is not king of the jungle; indeed what you notice most about tropical forests is that there=s very little life on the ground and therefore nothing to attract predators in off the savanna B whereas on the savanna they find great hordes of herbivores just asking to be devoured. Indeed, there were many more predators (and more different species of predators) around two million years ago than there are today, including some much bigger and more fearsome than those of today. Our ancestors of two million years ago, however, were much smaller than we are. Yet, just like us, they lacked the natural offensive weapons B claws, big sharp teeth, high speed over short distances B that are the weapons of choice for savanna predators. A species so ill-equipped would soon have gone extinct if its members hadn=t devoted far more time to predator detection and predator avoidance than apes do. But predator detection and predator avoidance depend not on social intelligence but on what we might call Apragmatic intelligence@: the noting and interpretation of clues in the environment (footprints, crushed vegetation, and so on), something that apes don=t seem to have.

Then there=s the question of food. Unless you=re adapted to eating grass B one of the few things that hominids weren=t adapted to eat B there really isn=t much food in savanna country, and what little there is comes in a bewildering variety of forms. There are a few trees and bushes here and there that yield fruit or nuts or berries, but because these are widely scattered and only bear for short periods at different times of the year, you have to remember what=s in fruit, where and when, if you=re not going to have many wasted journeys. You can=t just wander around thinking you=re bound to stumble on something sooner or later. There are roots that you can dig up and eat if you know what the plants growing out the top look like and whether they are poisonous or not. There are eggs or fledglings in the nests of ground-nesting or shrub-nesting birds, which are usually well hidden and which, apart from stumbling on them, can only be located by carefully watching and correctly interpreting the behavior of parent birds. There is the honey in the nests of wild bees that is delicious and highly nutritious, but extremely hard to get at without painful and possibly fatal consequences. There are small terrestrial animals, particularly the young of these, which, if you are skilled in observing and interpreting the traces they leave, you can sometimes kill with a thrown stone or by running them down. There are fish in the rivers and pools which, even before the invention of hooks and nets, you could, with immense caution, creep up on and catch by hand (at least, one of my wife=s uncles could, and surely any non-tool-using foraging skill accessible to effete modern humans must have been within the reach of our remote ancestors). There are the carcasses of dead animals of varying size, locating of which is easy (especially if you are watching the vultures) but which immediately puts you into direct competition with other, more powerful and better equipped scavengers. (Remember that the old dichotomy between scavengers and predators has been broken down by recent studies B any predator will scavenge if the bodies are there, just as any scavenger will predate if the prey can be caught fairly easily.) All these many and varied food sources were not conveniently located in an ape-sized terrain and continuously available, but scattered thinly over dozens of square miles of terrain and three hundred and sixty-five days of every year.

We tend to think of ourselves as continuations of modern apes (you can always get a feature article out of asking, are we more like the aggressive, opportunistically carnivorous chimpanzees or the less aggressive, more sensual bonobos?). Modern ethology has given us a picture of ape social groups spending many hours just lollygagging about, like modern humans at a holiday camp, teasing, grooming, play fighting, having a whale of a good time. It=s easy to imagine things have always been that way. We assume too often that the common ancestor of chimps, bonobos, and us was something pretty much like a modern chimp or bonobo, maybe halfway in between, and that the others more or less stood still while we changed radically.

---

WHC: The so-called competitive exclusion principle suggests that the extant apes might be the ones that avoided direct competition with our ancestors, for example, that the gorilla survived because it retreated into a vegetarian niche. Chimps and bonobos may not be perfect stand-ins for our common ancestor, but the many social behaviors that we share with them (see Frans de Waal=s books) suggest that we=re seeing ancestral behaviors, not ones separately evolved in the last five million years.

---

Unless you believe that ecology, the environment, and the way animals get food have no effect whatsoever on a species, you have to believe that in important ways, we were different long before we talked, maybe even before we walked upright. A lot of typical ape behavior would have had to be suppressed, or at least severely curtailed, under the conditions that early hominids experienced. Now that our niche allows them, now that our control over nature is complete enough to give at least some of us the necessary leisure, we can give free rein to those social behaviors, and indeed I=ve been told that the idle rich spend whole lives on them. But to say that because we do something now, and apes do it now, and our common ancestors did it, therefore at every stage of our evolution we must have done it B that=s saying the environment doesn=t matter in evolution, only genetic inheritance counts. I=m sorry, that=s only half the evolutionary equation. The environment does matter. Interaction between animals and environment is what evolution=s all about.

Social life didn=t get more complex for our remote ancestors; what did get more complex was the interaction between our ancestors and their environment. Increased wariness, increased curiosity, a far greater and more ruthless concentration on the exigencies of getting a living B these must have been the qualities that most sharply distinguished early hominid life, and that paid off most in terms of procreation and perpetuation of genes. Bonding, and the reciprocal altruism that cemented bonds, would of course have remained vital B you would want to be sure that, in a confrontation with predators, your buddy stayed by you or at least covered your retreat by throwing stones. But the strongest selective pressures would have come from the brute exigencies of survival. If we want to look for the pressure that first set language going, we must therefore look at the kind of life early hominids lived and the kind of behavior that this life would foster.

The foraging story goes as follows. Either the group foraged as a single unit or split into smaller units to forage. If they split, the story is straightforward. If one unit discovered an abundant food source that could feed all of them for a day or two, how would they tell the other units this? In other words, how could they evaluate competing food sources? That problem doesn=t go away even if the group foraged as a single unit. Remember that our ancestors had to be a lot more aware of their physical environment than arboreal apes.

Now suppose the group has just found a store of honey and is exploiting it, getting stung in the process. One member happens to look up and sees vultures circling just above the horizon. There=s dead meat there, maybe some dead megafauna that would feed the group for days. He jumps up and down, pointing. They=re too busy with the bees and the honey to take any notice. What=s the advantage to him if he can get their attention? If he=s right, and they make a big food find, he=ll be a hero, the others will look up to him, his status in the pecking order (and his access to mates) will be substantially enhanced. If he does this sort of thing consistently, he may get to be the leader. But how can he make them understand? If only he had a real word or two, something more specific than follow-me!

Now of course these are just-so stories, and we all know that just-so stories are just that: stories. What=s wrong with them, it=s claimed, is that you can make up a just-so story for any scenario you can think of. Or so they say. To date, nobody that I know of has produced a just-so story for the social intelligence theory of language origin. That is, nobody working within that particular framework has suggested what the first symbolic utterances might have meant, and under what circumstances they would have been uttered. The point here is that anything, even the crudest and most limited form of protolanguage consisting perhaps of a handful of words and/or gestures, has to have had an immediate payoff for the individuals who used it. If it didn=t, the behavior wouldn=t have continued, certainly wouldn=t have gotten fixed into the genotype. Nobody doubts that social uses of language (or protolanguage) would have paid off enormously in terms of individual fitness, but it seems to me that before they could do that B before our ancestors could say even simple things like AI like you@ or AYou got nice eyes,@ let alone AIf you and I fight together, we can lick Alpha,@ or AYour favorite female cheated on you with Beta this afternoon@ B linguistic ability would have already gone well past its starting point.

However, it=s easy to imagine what the first symbols might have been in either of the stories I told just now. In the first, something like AMammoth! Mammoth!@ (which could have been the noise the beast made, or miming a trunk, or anything that worked) plus ACome on!@, which might have been repeated arm movements in the direction of the circling vultures. In the second, something like ALeopard!@ (however that was rendered) plus pointing at the footprints and perhaps then raising a finger to the lips to indicate silence and caution. But so far, no supporter of the social-intelligence theory has come up with anything comparable. I=d be interested to see them try.

There=s one important corollary to the instruction story. One of the most interesting findings about Kanzi the bonobo is that, like human children, he engages in pretend play. Kanzi Aenjoys feeding a toy dog imaginary food@ or Amay pretend that a toy dog or toy gorilla is biting him@; he asks Sue Savage-Rumbaugh to pretend to be a monster and chase him, and pretends to be afraid of her, although it=s fairly obvious that he isn=t. It=s risky to generalize from one individual to a species, especially if that individual=s upbringing is as untypical of bonobos as was Kanzi=s. But it does raise the possibility that pretend play is genuinely homologous, and that accordingly we may suppose that the children of hominids engaged in it. If they did, then there can be little doubt that those children took the symbols they learned from their elders and incorporated them into their play, experimented with them, and expanded them. The relative plasticity of children=s brains and their capacity to both learn and innovate are not the be-all and end-all of the language story, but they may well have contributed significantly to it.

This chapter is, I know, largely speculative, and may never be proven or disproven. But nobody knows how much we may yet be able to learn about our remote ancestors. If we do learn more, it won=t be just a question of amassing facts. We can=t do without facts, but facts by themselves, alas, never say enough. All facts are subject to interpretation, and can take on completely different appearances depending on the lens through which they are viewed. Speculation forms a vital component in science; it helps to interpret facts and to guide future research, but if the research it has guided turns up things that it didn=t predict B even things quite incompatible with it B that=s par for the course. Provided you don=t turn the goal into a religion, you get more out of looking with a goal in mind than out of blind fishing expeditions. Speculation is light luggage, it=s easy to junk it and try again.

At present, we still simply don=t know enough about the Magic Moment of the first symbolic utterances to be able to do more than speculate. Short of cloning Homo erectus from bone-marrow found in frozen bogs, which is still science fiction but might not always be, we will never know for sure what its linguistic capacities were. In the unlikely event that an erectus gets cloned in my lifetime, I=d like to be the one that gets to give it language tests. I=d predict that around age two or less it could be taught protolanguage and that it would be more inventive than apes have been, but that attempts to teach it true human language would fail dismally. If it did learn syntax, a lot of this book would need re-writing. If it didn=t, and couldn=t learn, it would go a long way to confirming the picture presented here.

Meantime, one or two things indirectly support that picture. Most of the first fifty words any child learns are nouns, just like the AMammoth!@ and ALeopard!@ of our just-so stories. Not hello, goodbye, please, thank-you, or any of the expressions you might predict if language had arisen to cement social networks. Words like these are used very frequently in children=s presence and are often actively modeled for them at an early age. AMilk!@ ASay please Sally!@ AMilk!@ ANo, Sally, say, >Milk, please!=@ AMilk! Milk! Waaa-a-a-a-a-ah!@ How many conversations like that have we heard, or even participated in! Indeed there are societies in which almost the only verbal interaction adults have with children is to teach them the niceties of polite discourse, the right way of talking to aunts, cousins, grandparents, and so forth. None of this seems to make much difference to what children learn first, which is a set of labels they can attach to salient objects in the world around them (or more exactly, to their concepts of such objects).

So, I would argue that the deepest roots of language will be found to lie in extractive foraging or the instruction of children, rather than in social intelligence B important as that intelligence was as a prerequisite for language. To be motivated to communicate information, you have indeed to be able to think, in some form, AI know X; Y doesn=t know X; I should gain Z by telling Y about X.@ But that doesn=t mean that social intelligence was the pressure that pushed us, and us alone, out of the infinite host of alingual species into a minority of one.