The Third Chimpanzee: The Evolution and Future of the Human Animal

by Jared Diamond

  With no prospects of ever getting that time machine, how can we hope to trace speech origins? Two fast-growing bodies of knowledge may let us start to build bridges across that gulf between animal sounds and human language.

  First, we’ll look at the animal side of the language bridge. This approach draws on new studies of the vocalizations of wild animal: the cries, calls, grunts, and other sounds that animals use to communicate among themselves. We are beginning to sense how far animals have come toward inventing their own languages.

  Then we’ll examine the human side of the bridge. All existing human languages seem infinitely more advanced than animal sounds. Yet one set of human languages may hold clues to some early stages in the development of language.

  Listening to Vervets

  Birds sing. Dogs bark. Many of us have a good chance of hearing some calling animal on most days of our lives. Our understanding of animal sounds exploded when we developed new tools and techniques, such as high-quality recorders to capture animal sounds in the wild, electronic software to analyze subtle variations in the sounds, and methods of broadcasting calls back to animals to see how they’d react. Animal vocal communication, scientists are discovering, is much more language-like than anyone would have guessed half a century ago.

  The most sophisticated “animal language” yet studied is that of the vervet, a monkey about the size of a cat that is common in Africa. Wild vervets are equally at home in trees and on the ground of the savanna or the rain forest. Like other animals, they regularly face situations in which efficient communication would help them to survive.

  A female vervet monkey with young. Mother vervets give more alarm calls—warnings against leopards and other predators—when they are with their own offspring than when they are with other monkeys.

  About three-quarters of wild vervet deaths are caused by predators. If you’re a vervet, it’s essential to know the difference between a martial eagle, one of the leading killers of vervets, and a white-backed vulture, a bird of about the same size as the eagle but that eats dead things and is no danger to live monkeys. When an eagle appears, you must take defensive measures, and tell your relatives. If you fail to recognize the eagle, you die. If you fail to tell your relatives, they die, carrying some of your genes with them. If you think it’s an eagle when it’s really a vulture, you waste time running down from the treetop while other vervets are safely up there gathering food.

  Besides the problem of predators, vervets have complex social relationships with one another. They live in groups and compete for territory with other groups. If you’re a vervet, you need to know the difference between a monkey intruding from another group, an unrelated member of your own group who is likely to steal your food, and a close relative who will support you. If you get into trouble, you need to be able to tell your relatives that you are in trouble, not some other monkey. It would also be useful to share news about where to find edible fruits and seeds.

  Studies of vervet behavior have revealed that these monkeys communicate specific information about predators. When vervets encounter a leopard or other large wild cat, male monkeys give a loud series of barks, females give a high-pitched chirp, and all monkeys within earshot may run up a tree. The sight of an eagle overhead causes vervets to give a short, two-syllable cough, which causes other monkeys to look up into the air or run into a bush. And a monkey who spots a python or other dangerous snake gives a “chuttering” call, which leads other monkeys to stand on their hind legs and look down for the snake.

  These three calls by no means make up a vervet’s entire vocabulary. Other, fainter alarm calls appear to be triggered by baboons, jackals and hyenas, and humans. Vervets also utter gruntlike calls when interacting with one another. Even to scientists who have spent years studying vervets, the grunts all sound the same. Electronic measuring, though, showed differences in the grunts when monkeys interact with those higher or lower than themselves in their troop’s pecking order, or when they watch another monkey, or when they see a rival troop.

  How do we know the vervet’s calls are meant to communicate with other monkeys? Could they be simply automatic expressions of fear or alarm, the way a human might shriek with terror when watching a scary movie, even if no one else is around to hear? There are several reasons to think that vervets are deliberately communicating with one another.

  One piece of evidence is that a solitary vervet was observed being chased by a leopard for nearly an hour but remained silent through the whole ordeal. There were no other monkeys around, so it had no need to communicate. Another bit of evidence is that mother vervets give more alarm calls when they are with their offspring than when they are with unrelated monkeys. Finally, when two troops are fighting, a vervet in the losing troop may give the “leopard call” even though there is no leopard. The call sends all the monkeys scrambling into the trees. It is a deceptive “time out” that breaks up the fight. In addition, young vervets appear to learn how to utter and respond to sounds, just as human infants do. A young vervet’s “pronunciation” gets better as it gets older, and so does its reaction to the various calls.

  Are vervets sounds “words” or “sentences”? Does the “leopard call” mean “leopard” or does it mean “There goes a leopard” or “Watch out for that leopard!” or “Let’s run up a tree to get away from that leopard!”? It may be a combination of all of them. I was excited when my one-year-old son, Max, said, “Juice.” What I proudly saw as one of his first words, though, he meant as “Give me some juice!” Only at a later age did he add the syllables that show the difference between sentences and pure words. Vervets show no sign of having reached that stage.

  Apes That “Speak”

  Apes are more closely related to us than monkeys are, and they also make sounds. Vocal communication by wild chimps and other apes, however, is much harder to study than communication by wild vervets, because apes’ territories are much larger. The study of captive apes cannot reproduce all the qualities of an ape community in the wild. For this reason we are just beginning to learn about how wild apes use their natural “languages.” But through a different approach we have learned something about apes’ ability to communicate.

  Several groups of scientists have spent years training captive gorillas, common chimpanzees, and bonobos to understand and use artificial languages. The scientists have used plastic pieces of different sizes and colors as symbols to represent words. They have also used hand signs based on the signs deaf people use, and keyboards with each key bearing a symbol. Through these methods, apes have learned the meanings of as many as several hundred symbols. At least one captive bonobo, or pygmy chimp, has been found to understand a good deal of spoken English—but not to speak it. These studies show that apes have the mental ability to master large vocabularies.

  The anatomy of apes’ vocal tracts keeps them from producing as many vowels and consonants as we can utter. For this reason the vocabulary ofwild apes is unlikely to be anywhere near as large as our own. Still, I expect wild chimp and gorilla vocabularies to be larger than vervet vocabularies. Apes might use several dozen “words,” possibly including names for individual animals. In this exciting field in which new knowledge is being rapidly gained, we should keep an open mind on the size ofthe vocabulary gap between apes and humans.

  WHO BIT WHOM?

  HUMANS DON’T JUST HAVE VOCABULARIES of thousands of words with different meanings. We also combine those words, and change their forms, according to rules of grammar, including syntax, which means the rules of word order. Grammar lets us build a potentially unlimited number of sentences from a limited number of words.

  To appreciate the importance of syntax, think about the following two sentences. They have the same words but different word orders.

  “Your hungry dog bit my old mother’s leg.”

  “My hungry mother bit your old dog’s leg.”

  If human language did not involve grammatical rules, these two sentences would have exactly the same
meaning. Most linguists (people who study language and how it is structured) would not consider an animal’s sounds to be language, no matter how large its vocabulary, unless it also had grammatical rules.

  Do any animals have grammar? No hint of syntax has been discovered in vervet calls so far. Capuchin monkeys and gibbons do have calls they use only in certain combinations or sequences, but we have not yet figured out what this means. Some recent findings suggest that finches and possibly other birds may use a form of syntax in their calls, but more research is needed.

  Wild chimpanzees are the animals most likely to use grammar. I doubt, though, that anyone expects them to have evolved a grammar even remotely as complex as human grammar, which has prepositions; verb tenses to signal past, present, and future; and other structures. For now, it is an open question whether any animal has evolved syntax.

  The Human Side of the Language Bridge

  The gulf between animal and human communication is surely large, but scientists are gaining an understanding of how that gulf has been bridged from the animal side. Now let’s trace the bridge from the human side. We have already discovered complex animal “languages.” Do any truly primitive human languages still exist?

  To help us recognize what a primitive language might be like, let’s look at how normal human language differs from vervet vocalizations. One difference is that human languages have grammar, rules that govern how words are used to make sentences. A second difference is that vervet vocalizations stand for things that can be seen or actions that can be taken, such as “eagle” or “watch out for the eagle!” But up to half the words in human speech—words such as and, because, and should—do not refer to anything you can see or do.

  A third difference is that human languages have a hierarchical structure. This means that they consist of different levels: sounds, syllables, words, phrases, sentences. Each level is built on the level below it, and is bigger than that level. On the bottom level, we make a few dozen sounds. At the next level, these sounds combine into many different syllables. These syllables, in turn, combine to make thousands of words, which we string into phrases according to the rules of grammar. Phrases then interlock to make a huge, potentially infinite, number of sentences.

  The oldest known written languages date from about five thousand years ago. They were as complex as the languages of today, which means that human language must have reached its modern level of complexity long before that. Are there any peoples existing today with simple languages that might show us the early stages of language evolution?

  The answer is no. Some hunter-gatherers or other preindustrial groups in the modern world still use Stone Age tools, or did until recently, but their spoken languages are as modern and complex as ours, or as the written languages of fifty centuries ago. To investigate human language origins, we need a different approach.

  How New Languages Are Born

  One approach is to ask: Would people who never hear a fully evolved modern language invent a primitive language of their own?

  In the creole language of Guadeloupe, an island in the Caribbean, this sign reads, “Slow down, children are playing here!”

  In fact, children who grow up alone, away from other people, do not invent or discover a language of their own. But dozens of times in the modern world, whole populations of children have grown up hearing the adults around them speaking a drastically simplified language, similar to what children themselves speak around the age of two. These crude languages are called pidgins. Children who grow up hearing pidgin evolve their own language, far more advanced than vervet communications, more complex than pidgin, but still simpler than normal human languages. These second- generation invented languages are called creoles.

  Why would whole populations of adults talk like two-year-olds? Pidgins form when two groups who speak different languages need to communicate with each other—for example, when people from one part of the world start to colonize or trade in another territory. Within their own group, people speak their own native language. To communicate with people from the other group, they use pidgin. This is what happened when English-speaking traders and sailors arrived in the island of New Guinea in the early nineteenth century.

  At that time, the people of New Guinea spoke about seven hundred different languages. For both English speakers and the people of New Guinea to communicate across groups, there had to be a common, or shared, language that anyone could use. A pidgin developed, made up of simple words. In time that crude pidgin evolved into a more advanced creole that is now called Neo-Melanesian. Today Neo-Melanesian serves New Guinea not only as the language of much conversation but also of many schools, newspapers, and media and government activities.

  How does a pidgin become a creole? Let’s start with a look at how a pidgin works. Compared with normal languages, pidgins are poor in sounds, vocabulary, and syntax. An early-stage pidgin is mostly nouns, verbs, and adjectives. As for grammar, conversation usually consists of short strings of words, with few or no rules about word order or other variations. A pidgin is a kind of language free-for-all, with different speakers using it in different ways.

  If adults keep using their native languages most of the time, speaking pidgin only occasionally when necessary, the pidgin stays at the same crude level. But if a whole generation begins to use pidgin as its native language—for all social purposes, not just for work or trade— the pidgin will evolve into a creole, with a larger vocabulary and much more complex grammar.

  Even without an authority to set out the specific rules of this new grammar, the pidgin will become larger and more definite, until it is a creole. And although a creole is simpler than a normal language, it can express just about any thought that a normal language can express, while it is a struggle to say anything even slightly complex in a pidgin.

  HAWAIIAN CHILDREN CREATE A LANGUAGE

  IN THE LATE NINETEENTH CENTURY,

  American owners of Hawaiian sugarcane plantations brought in workers from China, the Philippines, Japan, Korea, Portugal, and Puerto Rico. Amid the chaos of languages, a pidgin based on English arose. Immigrant workers kept speaking their own native languages within their groups, while using the pidgin to communicate with people from other groups.

  Here’s an example of pidgin as it was spoken by people who came to Hawaii around 1900: “Me cape buy, me check make.” Cape, pronounced “kah-pee,” was the pidgin word for “coffee.” This example of a pidgin sentence could mean two very different things. It could mean either “He bought my coffee, he made me out a check,” or “I bought coffee, I made him out a check.” The people having the conversation would have to figure out the meaning depending on what else was being said or done at the time.

  Hawaiian workers did not improve on the pidgin they used, even though it was a very limited form of communication. That was a problem for the immigrants’ children born in Hawaii. Some children heard pidgin at home because their parents came from different ethnic groups. Even if both parents spoke the same native language at home, kids could not use that home language to communicate with children or adults from other groups. And social barriers prevented workers’ children from mixing with the plantation owners and, thereby, learning English.

  The kids’ solution was to turn the limited plantation pidgin into a full-fledged creole within a single generation. This process was recorded by a researcher who interviewed working-class Hawaiians in the 1970s. Because older people still spoke the form of language they had heard around them and learned in their youth, the researcher was able to trace steps in the transition from pidgin to creole. He found that creolization started around 1900 and was complete by 1920.

  Creole allowed young Hawaiians to express more complicated thoughts, using sentences that had a single, definite meaning. For example, “Da firs japani came ran away from japan come” meant “The first Japanese who arrived ran away from Japan to here.” And “One day had pleny dis mountain fish come down,” meant “One day there were a lot of these fish
from the mountains that came down [the river].”

  Hawaiian children created creole from pidgin as they learned to speak. While they communicated among themselves, grammar evolved. The result was something different from English and also different from the languages of all the worker groups. The children had invented a language.

  The Language Blueprint

  The birth of a pidgin that grows into a creole is a natural experiment in language evolution. It has unfolded dozens of times in the modern world, from South America through Africa to the Pacific islands, from at least the seventeenth century to the twentieth.

  The outcomes of all these language experiments are remarkably similar. Most creoles share certain features, such as putting a sentence’s subject, verb, and object in that order. It is as if you drew a dozen cards fifty times from well-shuffled decks and almost always ended up with no hearts or diamonds, but with one queen, a jack, and two aces. How could creoles that evolved in so many different places and times be so similar?

  I think the likeliest explanation is one that some language researchers have suggested. They think that humans share a genetic blueprint for learning language during childhood. In other words, much of language’s structure is hardwired, or programmed, into us by our genes. A hardwired language structure might produce the patterns that we see again and again in creole grammars. And on that foundation, over time, people could have built the great variety of the world’s fully developed languages.

  Now let’s pull together animal and human studies to try to form a picture of how our ancestors progressed from grunts to Shakespeare’s sonnets. The first stage is animal calls that carry specific meanings, such as those of vervets. The single words of young human toddlers represent the next stage—not just a grunt, but a word assembled from a set of vowels and consonants that can also be used to make many other words.