On the question of how language evolved, however, Chomsky was surprisingly incurious. Moreover, he shocked his scientific colleagues when he rejected the idea that language emerged through Darwinian natural selection.19 When pressed for an alternate explanation, he mused that language somehow appeared through some unexplained physiological fluke resulting from (as he once put it) packing so many billions of neurons into a space “the size of a basketball”20—an explanation only marginally more creditable than the one offered in Kubrick’s 2001: the visitation by the mysterious monolith and the great leap in intelligence that results. Still more surprisingly, Chomsky insisted that, however language appeared, it did so not for sharing ideas, plans, and goals through vocal communication. Language first appeared, Chomsky said, for the purpose of solitary, pristine, platonic—and perfectly silent—thought.21 Which ruled out, of course, any role, in the evolution of language, for the human voice.
Even some of Chomsky’s greatest champions have acknowledged that he has become a “guru” in the field of linguistics, with all the attendant dangers that go with slavish adherence to a single person’s ideas.22 Not every linguist, however, followed Chomsky’s dogma. The most notable apostate is Philip Lieberman, now an emeritus Professor of Linguistics at Brown University. For over fifty years, Lieberman has built his own theories of how language is learned and where it came from. In doing so, he has, virtually alone among serious scientists, proposed a central role for the voice.
Lieberman was one of the first students to enroll in Chomsky’s linguistics course at MIT, in the late 1950s, before Chomsky published his evisceration of Skinner’s Verbal Behavior and became famous. The class had just three other students.23 Lieberman was at first enthralled by Chomsky’s ideas, but soon came to question his conception of language as a purely mental phenomenon, a medium for thought. Clearly, Lieberman reasoned, language has an important physical dimension, as a spoken medium produced by the vocal organs for communication. Having earlier trained at MIT as an electrical engineer, Lieberman was, by temperament and training, fascinated by the physical apparatus that makes any complex mechanism function. Language was certainly a complex mechanism—as was speech, the primary means by which we transmit language. Lieberman wondered if clues to the nature and origins of language might not lie in the apparatus that makes language so useful to us: the voice.
For his PhD, Lieberman focused on the primary anatomical system that drives voice and speech—breathing—and performed groundbreaking work on an amazing, if underappreciated, aspect of our talent for talking: namely, how we gear the amount of air we inhale to the size of the thought we intend to speak, a lightning-fast calculation we perform depending on whether we plan to say “Pass the salt, please,” or “Four score and seven years ago our fathers brought forth on this continent, a new nation, conceived in Liberty, and dedicated to the proposition that all men are created equal.” This skill actually takes until puberty to master, as Lieberman showed by pointing to the speech of children as old as twelve who, despite clear articulation and considerable syntactic sophistication, sometimes find themselves running out of air midsentence.
Consulting respiratory physiologists, Lieberman learned how stretch receptor nerves in the rib cage’s intercostal muscles—the muscles between each rib that expand the rib cage when we inhale, and compress it when we exhale—help us gauge when the lungs have been inflated to the “correct” amount for a given utterance. (Similar stretch receptors in the bladder notify us that it’s time to find a bathroom; those in the stomach alert us to when we’ve eaten enough.) These stretch receptor nerves send signals from the muscles to the brain, but also in the opposite direction: from the brain to the muscles, creating a two-way circuit. Lieberman argued that this feedback loop in the nerves that control respiration, over the course of evolution, helped early humans to refine their breathing to fit the meanings they wanted to convey. We divide our thoughts into utterances that can comfortably be spoken on a single exhalation—which suggested, to Lieberman, an anatomical basis for one of the primary syntactic units: the sentence. We can even time our inhalations, he pointed out, around natural pauses within sentences, the places where, in writing, we would use commas to enclose an embedded clause (“… a new nation”—inhale—“conceived in Liberty…”). The fact that we shape grammatical structures around our breathing seemed to Lieberman clear evidence of how an “anatomical constraint” (the amount of air we can pull into our lungs) helped to create the structures of language.
This put Lieberman sharply at odds with Chomsky, who said that our uniquely human, grammar-generating “language organ” developed purely as a medium for thought, and is sequestered in the highest regions of the cortex. But breathing is controlled by older parts of the brain—such as the brainstem and cerebellum, structures we share with every vertebrate. Perhaps, Lieberman thought, grammar and syntax are not confined to highly specific areas of the cortex unique to human beings—a dedicated “language organ.” Perhaps it made more sense to think of language as distributed throughout the brain, in regions we share with our oldest nonhuman ancestors. In which case, Chomsky was surely wrong to say language could not have evolved through the normal processes of natural selection.
After graduating from MIT in 1966, Lieberman was hired as a full professor at the University of Connecticut and a member of the research staff at New York City’s Haskins Laboratories, the leading language research institution that was shortly to publish the seminal motor theory of speech perception. At Haskins, he continued his investigations into the bodily changes that might have driven the evolution of language. One logical place to look for the origins of speech, he thought, was in the sounds and vocal apparatus of our closest living animal relations: apes.
At zoos in Manhattan and the Bronx, Lieberman taped hundreds of hours of chimpanzee, rhesus monkey, and gorilla vocalizations and concluded that, while apes make some good consonants, especially lip sounds, like p and b, and the nasals, m and n, their vowels were limited to the schwa, the “uh” sound24—a limitation that he realized, through his familiarity with research from the 1950s on how we form vowels, was imposed by their high-placed larynx. Consulting Victor Negus’s classic Mechanism of the Larynx, Lieberman furthermore learned that we are the only animal with a permanently descended larynx, the result of the slow migration of the voice box down the throat that probably began when our ape ancestors stopped knuckle walking and stood upright. 25
Our weirdly low larynx had actually puzzled Darwin, because it brings the opening to our windpipe right alongside the opening of our esophagus, the tube that leads to the stomach—a dangerous arrangement, Darwin noted, because of “the strange fact that every particle of food and drink which we swallow has to pass over the orifice of the trachea, with some risk of falling into the lungs.”26 In fact, until the advent of the Heimlich maneuver in the 1970s, thousands of people died annually by choking on food that “went down the wrong way.” Darwin was understandably mystified as to why we evolved such an anatomical booby trap in defiance of everything he knew about natural selection, which is supposed to enhance our chances of living, not the reverse. Lieberman (in possession of knowledge that had not yet emerged in Darwin’s day about how vowel sounds are created by the blending of the overtone frequencies produced in both the mouth and the throat) concluded that the descent of the larynx conferred an important survival advantage that outweighed the very real danger of choking to death: a throat resonating chamber that allowed for articulate speech through a variety of clearly discernible vowels.
It thus occurred to Lieberman that if he could track, in the fossil record, the descent of the larynx from its high position in the chimp’s throat, to its uniquely low position in Homo sapiens, he might be able to create a rough timeline for when language emerged. The first hominid he studied was our closest extinct human relative: the Neanderthal, a human species that emerged around 500,000 years ago and which migrated out of Africa at least 200,000 years ago, populating large swat
hs of Europe and western Asia, before mysteriously going extinct about thirty thousand years ago—not long after encountering our species, Homo sapiens (who had eventually followed them out of Africa). Debate had long raged over whether Neanderthals could talk. They certainly seemed to have the necessary cognitive power (skull fossils show that their brain was slightly larger than ours). But some influential scientists insisted that language did not arrive until long after the appearance of Neanderthals—indeed, as recently as fifty thousand years ago, with the unexplained surge in cognitive power known as the Great Leap Forward. Chomsky endorsed the Great Leap theory. Lieberman didn’t buy it. He was increasingly convinced that speech and language were gradual developments that depended on the coevolution of many different voice-related anatomical and cognitive systems stretching back hundreds of thousands of years—or hundreds of millions, if you went all the way back to the lungfish (or indeed the ragworm). In keeping with this, he strongly suspected that Neanderthals could speak, and that the placement of their larynx might offer strong proof of it.
Using a cast of a Neanderthal skull discovered, during the early twentieth century, in the village of La Chapelle-aux-Saints, in France, and working with Edmund Crelin, the chief of anatomy at Yale New Haven Medical School, Lieberman determined that the Neanderthal larynx occupied an “intermediary” position between the chimp and human larynx. In other words, it had begun to migrate down the throat but had not descended to the level of ours. Crelin happened to be the first anatomist in the world to have investigated the anatomy of human newborns and had earlier revealed the chimplike high placement of the newborn larynx.27 Over the first few years of life, the larynx gradually descends. That the identical downward journey of the larynx occurred in our early human ancestors was wholly in keeping with an idea in evolutionary biology advanced by Darwin: namely, that “ontogeny recapitulates phylogeny”—or, in plain English, that the stages through which humans pass, from conception, as a single-celled zygote in the womb, to maturity, mirror the developmental steps our species took from single-celled eukaryote, in the primordial ocean, to our evolution, 2.7 billion years later, as Homo sapiens.
After all, as fetuses we are suspiciously fishlike aquatic creatures who use gill-like structures to extract oxygen from the liquid in our mother’s womb, and at birth we become air-breathing animals. In terms of our vocal tract, we are, effectively, chimpanzees who only gradually “evolve” into speech-capable humans as the larynx repeats the migration it made, after we diverged from Neandethals, almost to the bottom of the neck. Meanwhile, the sequential development of human vocal sounds parallels that of vertebrate brain evolution: a newborn’s first cries emerge from the (reptilian) brainstem as purely reflex reactions to pain or hunger; emotionally nuanced sounds appear around the third month as the baby’s mammalian limbic structures come online and its musical, prosodically expressive sounds help it forge social bonds with caregivers; only after that, when the larynx has descended sufficiently to allow the sculpting of rough vowel sounds, does the rational, executive cortex start to wire in, and the child begins to shape its emotional cries and whimpers, giggles and sighs, into the babbled phonemes and proto-linguistic “words” (like William’s mum) that the infant will eventually mold into articulate human speech.28
For Lieberman and Crelin, the burning question was how did the Neanderthal’s slightly higher larynx affect its voice? Crelin built elaborate reconstructions of the Neanderthal vocal tract from silicone, and they showed, using a computer program to model the vowels that the reconstructed Neanderthal airway could produce, that our closest extinct relative, owing to its elevated larynx, could produce only a restricted range of vowels, wider than the chimp’s schwa, but not including our ee, ahh, and oo sounds—those requiring the most drastic reshaping of the two vocal tract resonators with the tongue.29 One year later, MIT speech researcher Kenneth Stevens showed that the ideal proportions of throat resonator to mouth resonator, for clear vowels, is a 1:1 ratio, with the length of the mouth equal in length to the vertical section of the throat.30 These “ideal” proportions did not occur until the apelike snout of the Neanderthal face flattened in our species. For these purely anatomical reasons, Lieberman said, Neanderthal vowels would have been blurred, ambiguous—a handicap that made all the difference to the respective fates of Neanderthals and Homo sapiens.
That our two species were sufficiently close, biologically and socially, to have mated and reproduced is a staggering fact only recently revealed by DNA recovered, and sequenced, from Neanderthal bones. As it turns out, every human being whose lineage can be traced to Europe or Asia (where Homo sapiens and Neanderthals briefly coexisted) carries 1 to 4 percent Neanderthal DNA; people whose direct ancestry is African carry almost none31 (for the simple reason that Homo sapiens and Neanderthals didn’t meet up—and hook up—until both had left that continent). So, we are, in every sense of the term, closer to Neanderthals than anyone had ever imagined. But it was this very closeness that, according to the merciless laws of natural selection, put our two species on a collision course. Lieberman cited Darwin’s claim that “each new variety or species, during the process of its formation, will generally press hardest on its nearest kindred, and will tend to exterminate them.” 32 Lieberman concluded that Neanderthal extinction “was due to the competition of modern human beings who were better adapted for speech and language.”33
That is, we had better vowels.
In 1984, Lieberman published The Biology and Evolution of Language, the first book-length study on language origins since the Paris ban of 1861. It included his research on respiration, the descent of the larynx, vowel production, Neanderthal extinction, and other evidence to argue that language emerged, roughly 400,000 to 200,000 years ago, through naturally selected improvements in the vocal apparatus, which fed back into the brain, creating the complex neural circuitry (including the “language centers” of Broca’s and Wernicke’s areas) that make speech possible. He denied the existence of an innate “language organ” and attributed our linguistic capacity to our species’ remarkable general intelligence—the same smarts that made us discover fire, build sophisticated tools, make protective clothing, and cooperate in large groups. The phenomenal speed of language acquisition in children was owing not to an inborn language instinct (Chomsky’s Universal Grammar) but to Darwin’s “instinct to learn.”
Lieberman’s book was, by any measure, the strongest challenge yet mounted to Chomsky, but in a field dominated by the theory of linguistic innateness and Universal Grammar, it was almost completely overlooked by the linguistic mainstream. Lieberman nevertheless continued his investigations into how changes to the speech apparatus gave rise to language, including research into an area that, because of its purely speculative nature, he touched on only lightly in The Biology and Evolution of Language: that our linguistic capability could only have come to full fruition (after refinements to our breathing and the descent of the larynx) with evolutionary changes to those parts of the nervous system that make possible the incredibly rapid, exquisitely well-coordinated, and meticulously sequenced movements of our tongue, lips, and larynx for speech. (To produce even as seemingly simple a sentence as “Pass the salt, please” requires the coordination and careful sequencing, in under one second, of hundreds of muscles throughout the trunk, larynx, tongue, lips and face.) Lieberman had included some preliminary findings about this in his book,34 but few at the time were willing to credit the lowly motor system (parts of which we share with earthworms and mollusks) with a major role in the evolution of a higher-order process like language, which is, after all, an expression of thought. What possible connection could there be between thinking and the parts of the brain that control bodily movement? But since then, researchers have pinpointed cell death in the brain’s motor centers as the cause, not only of Parkinson’s patients’ tremors and speech impairments (they mix up voiced and unvoiced sounds), but of disturbances in thinking known as Parkinson’s Dementia. Lieberman’s own study, i
n the early 1990s, of mountain climbers suffering from the temporary dementia of high-altitude sickness from oxygen starvation of the brain also showed the link between higher thinking, speech, and crucial movement centers in the brain.35 The most dramatic evidence, however, was yet to come.
* * *
Many of the most momentous breakthroughs in science begin with the study of anomalies and abnormalities of human behavior—and such was the case for the evidence that began to emerge, in the late 1990s, in support of Lieberman’s intuitions about the critical role, for language, in evolutionary changes in the motor control of our larynx, lips, and tongue. The anomaly, in this instance, was an unfortunate West London family with an extremely rare speech disorder that afflicted several family members over three generations (mother, children, and grandchildren) and included an inability to pluralize words by adding s, and confusion over how to make verbs into the past tense by adding –ed.36 This runs-in-the-family, genetically determined disorder was initially seen as clinching evidence that our grammar and syntax must indeed be inborn, just as Chomsky said, and that every aspect of language might be controlled by specific genes that build particular parts of the cortex, tiny regions of the brain dedicated to, say, inserting subordinate clauses, or turning statements into questions, or making a verb into the past tense. Grammar genes!37
But this interpretation was soon ruled out in a groundbreaking paper of the mid-1990s by a team of London researchers who had been studying the family for a decade and who revealed that the afflicted family members’ linguistic difficulties were intimately tied to a mysterious “palsy” or paralysis that limited the movements of their tongue and lips.38 In 1998, the team published brain scans that revealed the source of this palsy: all the afflicted family members had an abnormal, underdeveloped area of the brain that controls fine motor movements: the basal ganglia, that ancient part of the brain that, as we saw earlier, plays such an important role in grooving-in the complicated, highly coordinated lip and tongue movements when a baby is learning to speak (and whose activity is shut down by oxygen starvation in high-altitude mountain climbers, and by cell death in Parkinson’s disease). The family members with no voice and speech problems had normal basal ganglia; the afflicted family members had basal ganglia that was noticeably shriveled and underdeveloped.39
This Is the Voice Page 12