The First Word: The Search for the Origins of Language

by Christine Kenneally

  Dan Sperber, a social and cognitive scientist at the French Centre National de la Recherche Scientifique in Paris, makes an interesting case for a component of language that probably predates fully modern language but must have evolved after our ancestors split from chimpanzees. Sperber is well known for the theory he and the linguist Deirdre Wilson presented in a seminal 1986 book, Relevance: Communication and Cognition. Briefly, relevance theory holds that inference is as fundamental to linguistic communication as the ability to decode the words in a given utterance. For example, the sentence “It is too slow” may convey a variety of completely different meanings, given different contexts. Sperber lists a few of the possibilities: “The mouse is too slow in solving the maze; The chemical reaction is too slow compared to what we expected; The decrease in unemployment is too slow to avoid social unrest; Jacques’ car is too slow (and so I’d suggest we take Pierre’s).”4

  Human communication in this view is about one person indicating his or her meaning to another. This can be done in a number of ways—via gesture, pantomime, or a linguistic code. The fundamental principle is that the person doing the listening (or watching, etc.) infers the speaker’s meaning from the signal and the context in which it is conveyed. (The relevance of any message results from a shared set of assumptions. It is crucial, for example, that the listener knows that the speaker wants to convey a meaning, also that the speaker knows that the listener knows this; in addition, there is a shared assumption that what is communicated, regardless of the form it takes, makes sense within the context of the communication.)

  Sperber and Wilson’s theory effectively crystallized the intuition that the context of language really matters, and since then, depending on their focus, researchers have placed differing emphasis on the relative significance of the pragmatic aspect of an utterance versus its linguistic structure. Regardless of these differences, Sperber makes the point that all the linguistic sophistication in the world won’t make language useful if its users are unable to infer the intentions behind an utterance and appropriately judge the relevance of its context. Likewise, Sperber points out that, compared to humans, chimpanzees have only a rudimentary ability to make inference about the beliefs and intentions of another chimpanzee. The ability to make sophisticated inferences about the relevance of a signal must therefore have preceded the final elaboration of structure in modern language, and it probably came after the split of our lineage with that of chimpanzees.5

  It’s not yet clear what type of investigation, experimental or otherwise, may further illuminate the relationship between pragmatics and linguistics in the evolution of language. Nevertheless, Sperber’s broad point is that both must be explored. In addition to the ideas put forth by Terrence Deacon and Michael Arbib in chapter 14, he has offered an excellent candidate for a specifically human precursor to modern language.

  If recursion did not transform our six-million-year-old grandparents into modern humans, perhaps it changed ancient humans into us by converting an archaic, simpler language into the version we have today. Is it possible that this is what happened two hundred thousand years ago? Yes. But if so, this shift is only one of many important turning points in the course of language evolution.

  Ultimately, the notion that a single attribute will explain why humans are the only living species to have language is as unhelpful in its latest version as in its oldest. There are hundreds of gateways to linguistic communication, and the evolutionary process provides no motivation to hail one, distinct from all the others, as more integral to language. The problem “Is there one crucial gateway to language through which only modern humans have passed?” may still be much discussed, but in all of its forms, it is truly a nonquestion.

  If good science doesn’t focus on one stage in linguistic evolution at the expense of all others, it will inevitably highlight only a selection. This is because some steps will be more experimentally tractable, while others will be easier to observe. Some steps will be notable because they preceded or followed a dramatic cultural shift.6 Some may be considered research-worthy because of their extreme remoteness in time, because they result from a human-specific genetic mutation, or because they drove the selection of a relevant mutation. If a stage in language evolution were ever linked to one of the few genes unique to Homo sapiens, it would draw enormous interest. Naturally, some steps will just seem more interesting because of what we think they imply about us.

  In the current debate, even though different researchers talk in terms of continuity and discontinuity or qualitative versus quantitative differences, there is nevertheless a greater and more important convergence on the same data and some basic concepts. To a large extent, the conflicts noted here are characterized by different emphases and focus rather than by completely opposed ideas. The argument between Pinker and Chomsky and their coauthors about FOXP2 illustrates this rather well. Pinker and Jackendoff argued that the importance of FOXP2 is that its sequence is uniquely human. Chomsky, Hauser, and Fitch discuss FOXP2 in very different terms—the gene that subserves language is shared by many different species and is therefore likely evidence of the broad foundations of language.

  Both are right. The shared nature of the gene implies an ancient history and widely dispersed potential for development along the language path. Nevertheless, a uniquely human mutation of FOXP2 has been positively selected in our species within the last 200,000 years. The FOXP2 mutation is a significant twitch on the genetic dial that accompanies the emergence of human language. Beyond this, the individual researcher may decide what matters most to him—the dial or the twitch.

  As for whether there is anything unique to language in the human brain, the question becomes complicated by the need to consider the development of the individual, the development of the species, the way that language itself changes through time, and the way that all of these factors interact. What’s certain is that the question no longer makes sense in the terms in which it used to be posed, that is: Is there a specific gene that programs a specific chunk of the brain to be a language processor? Nevertheless, it does appear that language doesn’t just fall out of the adult human brain without some specifically linguistic processes occurring, as shown by the neural route taken by regular past-tense verbs in contrast to irregular ones.

  What about the evolutionary processes of adaptation, where a trait evolves for a particular purpose, and exaptation, where a trait that is used for one function becomes co-opted to serve another purpose in later generations? What role have these played in language evolution? For all the furious words expended on the subject, everyone agrees that both processes have had a role. And everyone has acknowledged that communication has to have something to do with language evolution. Regardless, the rapid spread of the human mutation of the FOXP2 gene is definitive evidence that there has been positive selection for a form of gene that had major consequences for language.

  It’s not just the genetics that make the spandrel suggestion unlikely. Humans accumulate a great deal of knowledge in their lifetimes. They are also an extremely social species. Could it just be a coincidence that we are able to communicate all that knowledge to other humans? “We’re social creatures,” said Pinker. “We don’t just cooperate with our kin, we negotiate agreements with people that we’re not related to, and societies are formed by implicit social contracts and exchange and understanding. If language was really just a by-product, one wonders why there would be such an amazingly good fit into the rest of what makes us zoologically unique.”

  Many exciting angles remain to be further explored—for example, what’s essential to language development and what is helpful but ultimately incidental? Language clearly bootstraps itself from gesture, but does a species have to have gesture to develop some form of language? How many individuals do you need in a species, as well as in a community, for language to arise in the first place and for it to be passed down through the generations and keep evolving itself? If you have to be human to have human language, could another
species in different conditions ever evolve a form of language that used enough of the same basic building blocks that we could translate between our language and theirs?

  The jury is out on these questions, though there is every reason to believe that the more data that are generated, the closer we will be to an answer. We can expect resolution on how powerful an evolutionary force communication has been and what elements of language it has shaped. In addition, we can hope to know more about how fine-grained the back-and-forth of modification and selection has been. Were some spandrels adopted as a piece into language? Or did some small increase in the power to compute a grammatical relationship arise as a spandrel and then become further elaborated over a long history of adaptation?

  If there were a moral to the story of evolution, it would be that meaning is something that happens after the fact. There is no rhyme or reason to the mutations that occur over the evolution of a species. Within the constraints of what has so far developed, genetic mutations are random; it is what the creature does with them that makes them meaningful. Evolution is the opposite of destiny, and because we are creatures of both biological and cultural evolution, where we are going is really obvious only in hindsight.

  Certainly, it’s impossible to predict the future of Chomskyan influence. Chomsky is most famous in cognitive science for being the first to point out that language is both extremely complicated and innate. Now the main complications are how language is defined, what the goals of scientific endeavor are, and the strange and enormous sociological phenomenon that Chomsky has engendered.

  Within the field of language evolution, Chomsky is associated with the caveat that language may have as much to do with inner speech as it does with communication between two individuals. But the value of this caution is questionable. We all have the sense that words exist inside our heads and that this sensation accompanies thought. But what forms do the words in one’s mind take? How complete or incomplete are mental sentences? How could so subjective an experience even begin to be measured? No researchers have been inclined, or able, to take the basic idea any further than the form in which Chomsky first suggested it.

  Indeed, though Chomsky has thrown out this possibility on a number of occasions—and although he is interpreted by many as saying that this is why language evolved—on other occasions he has qualified it further. In 2000 he wrote, “One can devise equally meritorious (that is, equally pointless) tales of advantage conferred by a small series of mutations that facilitated planning and clarification of thought…not that I am proposing this or any other story.”7

  Chomsky’s focus on extraorganic principles and the idea that we just don’t know what happens when you pack that many neurons into a space that size is an important part of the debate. Recently a number of mathematicians at the Cold Spring Harbor Laboratory in New York announced that they had worked out the mathematical basis for why the brain is divided into white matter and gray matter.8 This is exactly the kind of idea that Chomsky has been promoting since the 1970s.

  Still, the way Chomsky carved up the linguistic universe is unacceptable to many researchers in language evolution. Even Ray Jackendoff proposes dismantling the long-standing Chomskyan ideas that the complexity of language arises out of the complexity of syntax and that syntax is central to language. Many researchers over the years took extremely seriously the idea that syntax was autonomous and somehow preexisted everything else in language. Gallons of ink have been spilled in the attempt to build models of a language processor that contains a separate syntactic processor, which analyzes the abstract structure of spoken language even before the sound. But, says Jackendoff, it’s time for this to be discarded.

  Typically, Chomsky has been ambiguous, enlightening, and dismissive of the new ideas about the emergence of language. For example, Terrence Deacon’s book The Symbolic Species was received with admiration by many in the field. Chomsky, on the other hand, wrote, “I have no idea what this means.” Deacon’s account of linguistics, according to Chomsky, is “unrecognizable.” He concluded: “I do not recommend this course either; in fact could not, because I do not understand it.”9

  One striking effect of the paper that Chomsky co-wrote with Hauser and Fitch was that it seemed to make other researchers in the field even more sensitive to, and critical of, Chomsky’s vast influence. Derek Bickerton, who years before had written that nothing really happened in linguistics before Chomsky, wrote about the Stony Brook conference on his blog:

  On October 14, 2005, Chomsky disembarked on Long Island for one of the few conferences he has attended in the last several decades: the Morris Symposium on the Evolution of Language at S.U.N.Y., Stony Brook. He arrived too late for any of the presentations given by other scholars on that date, gave his public lecture, gave his conference presentation at the commencement of the next morning’s session, and, despite the fact that all of the morning’s speakers and commentators were expected to show up for a general discussion at the end of that session, left immediately for the ferry back without having attended a single talk by another speaker. For me, and for numerous others who attended the symposium, this showed a lack of respect for everyone involved. It spelled out in unmistakable terms his indifference to anything anyone else might say or think and his unshakable certainty that, since he was manifestly right, it would be a waste of time to interact with any of the hoi polloi in the muddy trenches of language evolution.10

  Does the fact that Chomsky is now contributing to the discussion on language evolution mean that he is conceding it is crucial to linguistics? Pinker said no. “He gives with one hand and takes with the other. Chomsky says, ‘All hypotheses are worthless, so here’s mine, which is as worthless as anyone else’s.’” This latest gyration in a long career of twists and turns, Pinker said, marks the beginning of Chomsky’s decadent phase.

  What does it mean that one man had such a long-standing and wide-reaching impact? “I don’t think it is good,” said Pinker.

  Because Chomsky has such an outsize influence in the field of linguistics, when he has an intuition as to what a theory ought to look like, an army of people go out and reanalyze everything to conform to that intuition. To have a whole field turn on its heels every time one person wakes up with a revelation can’t be healthy. It leads to a lack of cumulativeness, and an unhealthy fractiousness. It’s an Orwellian situation where today Oceania is the ally and Eurasia is the enemy, and tomorrow it’s the other way around. Time and effort and emotional effort get wasted.

  Ray Jackendoff likened Chomsky’s persona and influence to that of Freud in psychoanalysis. “Freud especially is an interesting model,” Jackendoff said. “Even though the specifics of the way Freud thought about things have been shown to be incorrect, nowadays we still take for granted all the basic ideas of Freud’s approach to the mind, about people’s motives and what drives them. Everyone who goes to a therapist now owes it to Freud. The same is true of Chomsky. The idea that you can look at language as a computational system invested in the mind and that there’s an acquisition problem that requires some question about what the child is bringing to the learning process, and that there are formal tools for discovering language in great detail—that’s now taken totally for granted in the field and that came from him.”

  The study of language evolution is in some ways the opposite of the formal linguistics that Chomsky created. It doesn’t start with language as a formal abstraction, but grounds it first in the human body, and in history. The questions that Chomsky considers critical, such as “Is language useless but perfect or useful and imperfect?” are not much discussed outside considerations of his own work. As for the notion that linguistics poses a crisis for biology, most evolutionary biologists and other researchers in the field seem confident that they can be brought into consilience.11

  The power that Chomsky has wielded and still does is impressive. Many researchers regard the ideas in the Science language evolution paper as just the natural maturation and progression of a brillian
t mind. This one man and his unique ideas have influenced literally thousands of academics. In the early days of language evolution, his name was used as an obsessive touchstone in many articles. But people now seem to be freeing themselves from that influence.

  Few are up to the task of disentangling the ideas attributed to Chomsky from the ideas that really are his. Without a doubt, people hold him responsible for things he didn’t say. And he is often accused of denying things he did say.

  As for generative linguistics, in the gentle phrasing of Jim Hurford, it is taking the burden off universal grammar. Indeed, all the evidence about genes, gesture, speech, physiology, and brain damage point away from UG. Today, many researchers who argue that the innateness of language is neither language-specific nor grammatical in nature still use the term “UG.” Some researchers even go to the trouble of pointing out that what they mean by “UG” is neither universal nor a grammar, a caveat that surely qualifies the term as either misleading or irrelevant.

  Only time will tell if the magnitude of Chomsky’s influence will persist. Currently the divide between his many critics and supporters remains religious in its zeal, with many researchers believing that Chomsky is an academic villain who led linguistics completely astray. In some lights, however, their problem is a definitional one. Chomsky’s interest extends only to what he considers the syntactic core of language. This necessarily excludes all this other study. Why should this matter so much? Having interests, and therefore areas of indifference, is a freedom allowed most everyone else in academia, but Chomsky’s lack of interest in a topic often leads to umbrage. Others still see him as the source of everything we now know. Charles Yang, a professor of linguistics at the University of Pennsylvania and author of The Infinite Gift, wrote in the London Review of Books that Pinker and most other researchers are merely turning over the rocks at the base of the Chomskyan landslide.12