The First Word: The Search for the Origins of Language

by Christine Kenneally

  Chimpanzees, obviously, have basal ganglia. Birds have basal ganglia. So do rats. When rats carry out genetically preprogrammed sequences of grooming steps, they are using the basal ganglia. If their basal ganglia are damaged, then their separate grooming moves are left intact, but their ability to execute a sequence of them is disrupted. (Lieberman calls their grooming pattern UGG, universal grooming grammar.)

  The fact that a number of different animals use the basal ganglia for sequencing, whether it involves grooming or words, said Lieberman, suggests that there is no innately human specialization for simple syntax. Instead of being a contained and recent innovation in the human lineage, the foundation of syntactic ability is an adaptation of our motor system, a primitive part of our anatomy.

  Lieberman’s contrarian (at least prior to 1990) take on language and its history offers an entirely different way of thinking about language evolution. When he started engaging with the subject of language, he wrote of it as not so much a new thing that humans have as a new thing we do, and we do it with a collection of neural parts that has long been available to us. Moreover, when you think about language this way, it is not really a “thing” at all but a suite of abilities and predispositions, some recently evolved and some primitive. The many parts of the brain and body that make up the language suite allow us to program into our own heads how our parents speak. When Lieberman calls language part-primitive and part-derived, he echoes Charles Darwin, who wrote in The Descent of Man that language was half art and half instinct.

  The nineteenth-century German philosopher Arthur Schopenhauer said: “All truth passes through three stages. First, it is ridiculed. Second, it is violently opposed. Third, it is accepted as being self-evident.”

  The study of language evolution from the nineteenth century onward has rather neatly followed the same course as Schopenhauer’s aphorism. Linguists once considered pursuing the topic an absurd endeavor. Then it was banned. After that, the official ban developed fairly seamlessly into a virtual ban. Now, where most researchers once glibly proclaimed that you can’t study it, many say you can, including the scholar best known for saying you can’t (or at least, you shouldn’t bother).

  We are at a strange now-you-see-it-now-you-don’t moment in the history of language and mind where it seems that everyone is taking possession of the same new attitude. It’s remarkable, now that the rhetoric about language evolution has shifted, how quickly what was once heretical has become received wisdom. Within a few years, students in Linguistics 101 will probably assume that asking about language evolution was always this easy and obvious.

  In a relatively short time, academics like Savage-Rumbaugh, Lieberman, and Pinker, in their different ways, have had enough influence to make the subject no longer controversial or taboo but a legitimate line of inquiry—an endeavor about which reasonable people could disagree.6 When questioned about the investigation of language evolution at the 2005 Morris Symposium on the Evolution of Language, Chomsky himself shrugged his shoulders and said, “I wouldn’t have guessed it could go so far.”

  Of course, there are still profound disagreements among the researchers. Even though Chomsky published a paper that discussed language evolution in 2002, he remains immensely discouraging about the subject. In addition, he argues that it is possible to engage with language evolution for purely logical reasons that are internal to linguistic analysis. Pinker and Lieberman, on the other hand, build their respective cases with findings from genetics, psycholinguistic studies, and experiments that compare the cognition and communication of various animals and humans. However, they disagree completely about the nature of syntax. In 2003 Sue Savage-Rumbaugh announced that Kanzi had uttered his first spoken word, but of Chomsky, Pinker, and Lieberman, only Lieberman considers her work to have crucial insights for language evolution.

  Nevertheless, the findings of all these scientists are important touchstones, and thanks to their disagreements, engagement, and even disengagement, the field has widened considerably. A great multidisciplinary conversation is now taking place. American biologists, Italian physicists, Australian neuroscientists, British anthropologists, and a variety of linguists and computer scientists are but a few of the academics investigating the origins of language. Researchers like Marc Hauser and Tecumseh Fitch, who co-wrote Chomsky’s 2002 paper on language and evolution, are proposing an entirely new field of study—evolutionary linguistics. The consensus among these researchers resonates with one of Savage-Rumbaugh’s and Lieberman’s main points: that what evolved was not a single thing. Language is not a monolith.

  Part 4 of this book will delve more deeply into the disagreements provoked by the new questions, but first we will survey the diverse array of experiments that purport to explain what the components of language are and what processes were responsible for drawing them together. Lieberman showed that the basal ganglia are implicated in the evolutionary trajectory that led to language, but what other parts of the brain are involved? What do they contribute? What about thought? What about gesture and speech and words? How ancient are they? And what relationship do genes have with language?

  Michael Arbib, a neuroscientist who investigates mirror neurons, thought now to be important for language, prophetically claimed that as the field develops, “we are going to be dazzled and puzzled and infuriated by the number of ways that language is defined.”

  II. IF YOU HAVE HUMAN LANGUAGE…

  Rats execute one movement after another in a logical sequence when they groom themselves; it doesn’t mean they can tango or, for that matter, rap. Human linguistic ability, taken as a whole, is still completely unlike anything else in the biosphere—which is why most of the key skirmishes in language evolution revolve around the issue of uniqueness. Does language make us unique? Is language unique to humans? If rats possess the same sort of systems that we use for language, does it make us less unique? Does it make us more like rats?

  These are reasonable questions. Obvious and enormous distinctions exist between human life and other animal life on this planet. We communicate with animals, but we don’t converse with them, and they don’t talk to each other. They don’t read magazines, they don’t write books, and they don’t compose poetry. Language appears to be uniquely unique. Naturally, we want to explain why this is so.

  But asking what makes humans unique is almost always qualitatively different from asking what makes the antelope unique, or the sloth, or the dung beetle. These questions don’t have to be different, but they have historically been so. The former is never purely scientific, but is inevitably shaded by our self-regard and is always, to some degree, existential. We think that working out what distinguishes us from them will entail working out what makes us us.

  Throughout history, many “uniquely human” attributes have been proposed. We make tools. We are creative. We have culture. We play. Inevitably, these characteristics get disputed. Chimpanzees use tools, crows design and build them, half a dozen species have culture, practically everything plays. Nevertheless, we keep trying to draw an absolute line in the sand.

  Anthropocentrism crops up at every level of research. Science as a body reeled when it was announced in 2001 that the human genome consists of maybe 30,000 to 40,000 genes, only 10,000 to 20,000 more than a roundworm’s. It was always assumed that if organisms as simple as worms had about 19,500 genes, we must have many times more—at least 120,000 of them. Since the 2001 announcement, the estimate has dropped even further, and the same group that put the human genome at 30,000 to 40,000 genes revised the number downward in 2004 to 20,000 to 25,000.1

  For the same reasons that the size of the human genome was wildly overestimated for years, animal cognition has been prone to underestimation. Indeed, it is only very recently that the word “cognition” has even been used to describe animal thinking, planning, memory, and knowledge. Frans de Waal draws a parallel between the study of animal cognition and the topic of animal emotion: “Emotion is basically taboo still. For example, there�
�s a scientist I know who was one of the pioneers of empathy studies in children, and she told me that twenty-five years ago, if she presented anything on empathy in children, she would be categorized with groups like people who study telepathy. I think in animal studies we are now at that stage where she was twenty-five years ago.” He adds, “If I mention emotions in animals, a lot of scientists become very squeamish.2 So that’s still a taboo topic, and it has the same flavor as the cognition topic. There used to be objections to cognition studies, because people would say, ‘You cannot look into the head of an animal.’ You also cannot look into the head of a human.”

  De Waal’s book The Ape and the Sushi Master (2001) is an insightful account of the history of anthropomorphism and anthropocentrism in animal research, which he believes is still relatively widespread in science. In 2002 he said:

  Well, there’s still a large category of scientists who would strongly object to comparisons between humans and other animals. They don’t just want to keep humans and animals separate; they also want to keep them separate in the language that they use. It’s almost as if the world is divided into two kinds of people. There are certain people who object to the comparison with animals and feel insulted by it, and there’s another group of people who think it’s fine and, “What’s the big deal?” You see that in the sciences, but you also see it in philosophy and you see it in authors who write. You see it everywhere.

  Like studies of cognition and emotion, language research is one of the last areas in which ideas about what makes us biologically unique and what makes us personally special still get muddled. It’s not unusual to run across a statement like the following: “One of man’s greatest achievements is language”—as if all the species took an exam and humanity was the only one to pass. (No one talks about the grand successes of the fruit bat or the accomplishments of the pygmy owl.) It’s not hard to infer from statements like this that we had some kind of conscious control over the evolution of words and rules. But there is no agency in evolution; it is inadvertent. We survived, modified, and multiplied, just like any animal alive today, and out of the wildly dodgem course we took, language arose.

  The intimacy of our relationship with chimpanzees has been apparent from the fossil record for some time, but in recent years it’s been underscored by genetic evidence. It’s now a much-quoted truism that our DNA sequence is, on average, 98 percent the same as that of chimpanzees (and only marginally less than that with other primates). Defining what it is to be human has, for many researchers, meant looking only at the 2 percent difference and assuming that in that small gap lies the key to all the cities of the world, all the farms, the oil rigs, the fast-food franchises, all of culture, and language.

  This plays out in many different ways. For example, scientists assumed for a long time that the parts of the brain that have to do with language must be wholly new, recently evolved additions that we do not share with nonhumans. The underlying idea here is that we attained a superior enough level of intelligence or complexity to acquire language, a sort of evolutionary benediction.

  Because researchers who take this approach see little in common between human and nonhuman communication, investigating the linguistic overlap between us and other animals has traditionally been dismissed as wrongheaded or irrelevant. In this view, the implication also lurks that human language can be acquired as long as you have a sufficiently powerful brain, as opposed to a specifically human brain shaped by specifically human constraints. But if you could somehow make the brain of, say, a crocodile smarter in some general sense, it would not automatically burst forth with human language.3

  So if you happen to have human language, it means that you are…human. You are terrestrial, you are a mammal, you are bipedal, you are a social primate. And if you are human, your ancestors traveled a particular evolutionary path, and many animals alive today have ancestors that walked much of the same path.4 In this zoomed-out view of humans as animals, discounting the communicative abilities of other animals, argue scholars like Lieberman, is at best hermetic; at worst, it’s unscientific. Which of these traits are necessary for human language and which are incidental can be determined only if they are all considered in the first place.

  Although we have many ancestors in common with other species, we have more recently evolved alone. So how do you explain both continuity and uniqueness? The scholars who are trying to break down the monolith of language and work out where the seams lie seek the essence of humanity and of language in both the 2 percent and the 98 percent. Yes, language as we know it today is uniquely human, but, says Lieberman, “human linguistic ability can be traced to the motor response of mollusks.”

  Evolutionary biology takes the view that all animal kind, despite its gorgeous multiplicity, is merely a set of variations on a theme: life. You are cousin to the world’s roaches, puppies, and Tasmanian devils because you share an ancestor, that once-upon-a-time cell that winked, split, and got the whole thing rolling.

  The continuity of animal life is not just an intellectual orientation or frame of analysis; it’s a visceral, involuntary experience. In Leipzig, Germany, in 2004, a group of scientists got together to discuss gesture and language evolution. In one presentation, baboon communication was discussed. When you are a male baboon, communication can be very high stakes: it’s not uncommon for one of these animals to reach forward and rip the testicles off its interlocutor.

  When the presenter described this unpleasant form of exchange, a flush spread through the conference audience. Everyone present was a scientist of some sort, most worked with animals, yet the joking and laughter and the slight sense of hysteria that followed was semi-involuntary. The remark would not have had quite the same intensity if the animal in question was a bug and the body part an antenna. Human life is continuous with all life, and baboons are pretty close to us on the continuum.

  One way biologists measure traits is by asking whether they are shared with other animals. If a trait is shared by two or more species, the next question is whether the relationship is one of homology or analogy. If the traits shared by two or more species are homologs, then they have come to those species from a common ancestor, perhaps an ancient one, which also had the same trait. Human arms and bat wings are homologous. And you can cut the distinction finer: if a trait is homologous between two species, it’s useful to ask if it is effectively the same or whether there is only a partial similarity.

  If a trait is an analog, then even though it is shared by different species, the trait evolved separately in those species. Biologists say that these species have converged upon a similar solution to a problem. Bat wings and butterfly wings are analogs. A trait that is analogous between species may or may not be especially old in either one or both of them.

  Analogy is a useful concept because it demonstrates that it’s possible for the same trait to evolve even in very different species. For example, both humans and songbirds are exceptional at learning vocally—humans with language and song, and songbirds with the songs they grow up hearing. Very few species can match their skills in this respect. One useful observation that results from comparing the two is that you don’t have to be human to develop a particular expertise—even if that expertise is part of what makes us human.

  It’s important to be cautious when labeling a trait homologous or analogous. For example, humans think of themselves as tool users par excellence. But other great apes, such as chimpanzees, gorillas, and orangutans, are also skillful with tools. The mother of all the great apes, the ancestor of orangutans, gorillas, humans, chimpanzees, and bonobos, lived about fourteen million years ago. Over time, the descendants of this grand-primate diverged into two species: one that evolved into the modern orangutan, and one that was the common ancestor of gorillas, humans, chimpanzees, and bonobos. This latter creature lived for around seven million years before its line also split into two paths, one of which led to the modern gorilla and one that led to humans, chimpanzees, and bonobos. The common ancestor
of humans, chimpanzees, and bonobos lived about six million years ago, after which the line that led to humans split away. The human lineage threw up at least twenty different types of hominids, only one of which survives today, Homo sapiens sapiens. The other line split again a million or so years later and led to the two chimpanzee species that exist today, Pan troglodytes, the common chimpanzee, and Pan paniscus, the bonobo, found only in Zaire.

  The branching of the great ape family tree means that chimpanzees and bonobos are our closest relatives on the planet. Even though to the human eye a bandy-legged chimp may look more like a gorilla than like a human, the chimp is actually more like us than it is like the gorilla. As noted by the evolutionary biologist Jared Diamond, humans are really just the third species of chimpanzee. Perhaps to a gorilla, we all look alike.

  Based on how closely related we are to chimpanzees, you might assume that they would be the best tool users of the other great apes, and although there is good tool use data for them (as well as gorillas and orangutans), that’s not their reputation. There is a saying among primate keepers, Heidi Lyn explained, that if you give a give a screwdriver to a chimp, it will throw it at someone. If you give a screwdriver to a gorilla, it will scratch itself. But if you give a screwdriver to an orangutan, it will let itself out of its cage.5

  If traits are not shared, then it usually means they are species-specific. Language, as in the whole suite, is clearly species-specific—it is not shared and has no homologs or analogs on the planet. But what happens when you disassemble the monolith? To what degree is this new function made up of old parts? And how old is old? Do some parts share an evolutionary trajectory? And how many of them are shared? Are they completely interdependent or somewhat independent of each other? How many, if any, are uniquely specific to humans?