Out on a Limb

Abstract

It may seem like unwarranted editorial privilege to be allowed the opportunity to reply to a review of one's own book, although I am assured that the invitation does not arise from my role as editor. My fantasy would be to actually review the book, but that would clearly be beyond the bounds of decency. In any event, if I am honest, I have to admit that Bradshaw's review is probably more positive and generous, and certainly more helpful, than my own review would have been. Bradshaw rightly notes an evangelical tendency, and there is perhaps a resemblance between the scientific literature and the Bible, in that one can usually find in both what one wants to find. There is therefore probably little point in debating here the finer details of whether other species have "theory of mind" or intentional control over their vocalisations, or whether chimpanzee gestures in the wild are learned; these are important issues, but they will have to be resolved empirically. The bottom line, I think, is that it has proven very much easier to teach apes to communicate using visuomanual techniques than to teach them to speak. Indeed, the first reasonably successful system was based on American Sign Language, although greater success is now claimed through use of a keyboard. For example Kanzi, the celebrated bonobo studied by Savage-Rumbaugh and her colleagues, communicates by pointing to symbols on a keyboard containing 256 symbols, and supplements this with his own invented gestures. This is still not language in the way that linguists like to define it, but it is a lot closer to language than Kanzi is able to achieve by vocalising. Perhaps we will one day be able to harness the vocalisations of chimpanzees or bonobos into a communication system that rivals these visuomanual ones, but I would bet against it. It is very likely that the vocal abilities of the common ancestor of humans, chimps, and bonobos lay closer to those of Kanzi than to our own. The vocal system has undergone considerable change in the course of hominin evolution, no doubt driven by the selective pressure for enhanced vocal communication-a pressure that does not seem to have operated in other primate species. The evidence I reviewed in the book strongly suggests that these changes evolved late, and may not have been complete until the emergence of our own species, Homo sapiens, some 170,000 years ago-and possibly even later, as I shall argue, with new evidence, below. This has led some to maintain that language itself evolved late, in a kind of Big Bang that distinguished our species from all others. But can something so complex and powerful as grammar really have evolved in such a punctuate fashion? Given that signed languages have all the grammatical sophistication of spoken languages, and that we readily resort to manual gesture when speech fails or is denied, it seems to me much more likely that language evolved gradually, with speech finally assuming the dominant role as the vocal system evolved. So why is there so much resistance to the idea that gestures may once have played the dominant role, as they do in Kanzi? One reason, I think, is that it simply doesn't feel right. As Bradshaw puts it, he just "cannot conceive of a troupe of rapidly signing hominins." The best antidote to this understandable failure of imagination is precisely to witness "a silent, animated, deaf community," and observe just how expressive and natural are their conversations. Moreover, we are gestural creatures, with excellent control over the arms, hands, and face, a well-developed capacity to make ourselves understood visually when speech fails, and a compulsive tendency to gesture with our hands while we talk. As the Haskins group has pointed out (sic) for years, speech itself is fundamentally gestural rather than auditory, and I argued in the book that speech may be just an extension of orofacial gestures that were initially primarily visual. We still lip-read, as the McGurk effect demonstrates, but facial gestures were increasingly swallowed into the mouth and throat, so that it was only through vocalisation that they became accessible. The critical question, as I see it (sic), is not whether there was a visual component, but just how far it developed before being overtaken by a predominantly vocal system. My guess is that human language was never purely a matter of signing, but was always a mixture of sight and sound. Even Kanzi vocalises insistently as he signs, but the linguistic message lies in the signs rather than the emotional sounds that accompany them. It took a very long time for vocalisation to get the upper hand (sic), so to speak. Given that I devoted a chapter in the book to the signed languages of the deaf, Bradshaw rightly raises the question of language in the blind. If my hypothesis is correct, then earlier hominins who were blind would have had much greater difficulty than they do today. But of course language does now function as a largely autonomous vocal system, and the main disadvantage suffered by the blind is in understanding how words relate to objects and actions that they cannot see. In teaching a blind child to speak, one must still resort to some kind of gesture, presumably through touch and manual guidance. It has been observed that blind people nevertheless gesture manually as they speak-even on the phone, I'm told-suggesting again that gesture is not far from the surface. Could it be argued that speech itself evolved to facilitate communication with the blind? Well, even I would not go that far, but the ability to communicate with the blind would certainly have been one of the secondary benefits. It is gratifying that Bradshaw should have raised the issue of lateralisation of function, especially as the more we disagree the better it is for the future of this journal. I have previously expressed alarm at the possibility that the topic may go into hibernation, as it did after the turn of the 20th century, but it may well be questions about its evolution that will keep the topic alive (Corballis, 2000). I argued in the book (and elsewhere) that sensory systems, motor apparatus, and corresponding brain functions are fundamentally bilaterally symmetrical because the environment in which we live and move is without systematic left-right bias. Other things being equal, a symmetrical system is better adapted to a symmetrical environment than is an asymmetrical one. Nevertheless bilateral symmetry is violated in some functions, notably those that give rise to handedness and the lateralised representation and control of language, leading to the question of what it was in our evolutionary history that created the initial nudge. Perhaps it was the asymmetry of the internal organs. Bradshaw's money is on right-hemisphere processes associated with emotion and spatial processing. In the book I plumped for left-hemispheric vocalisation, in part because it has been documented in the frog, suggesting that it may have very ancient roots. Vocalisation, moreover, is not constrained by the spatial environment, so there is no compelling reason why its control should preserve bilateral symmetry. Nevertheless, whatever it was that created the initial asymmetry also set the stage for a degree of asymmetry in other functions, including spatial functions, that might otherwise have benefited from symmetry. Bradshaw considers it odd that I should propose that language lateralisation might be linked to vocal asymmetry in the frog, because this suggests to him that the gestural stage is redundant. But the origins of lateralisation are not the same as the origins of language-I do not claim that frogs can talk. My argument is that lateralisation of language, and indeed handedness itself, may have emerged when vocalisation was incorporated into gestural language. Support for this, reviewed in the book, comes from the mirror-neuron system that maps the perception of grasping movements onto the production of those same movements. This system is bilateral in the monkey, but left-hemispheric in humans. In humans, moreover, the system appears to include Broca's area, supporting the view that it was the incorporation of the vocal element that lateralised it. In short, language may have gone from hand to mouth, while lateralisation went from mouth to hand. I have elaborated this argument in more detail in a forthcoming open-peer article that will allow others to enter the fray (Corballis, in press). The provocative claim that spatial awareness is linked to the temporal lobe, and not the posterior parietal lobe, comes from respected authors writing in a respected journal (Karnath, Ferber, & Himmelbach, 2001), and I suppose really should be taken up with them. Nevertheless I am sympathetic to Bradshaw's counterclaim for the more traditional parietal and frontal sites, and his gracious concession that the critical area might well be the right-hemisphere homologue of Wernicke's area, which after all does extend into the parietal lobe. We might even allow the occipital lobe as well, since Wilkins and Wakefield (1995) have identified the parietal-occipital-temporal junction (POT) as critical to the evolution of language, and it seems reasonable to me to suppose that the right-hemispheric bias was achieved precisely because the left POT became absorbed with communication at the expense of the right POT. Perhaps we should not neglect the frontal lobe either, since Broca's area on the left may have deprived the left frontal lobe of some of its spatial functions. As for the association between handedness and magical ideation, it may well be implausible, but it rests on an empirical result (Barnett & Corballis, 2000) that also left me somewhat bemused. In any event, it perhaps signals another direction that may help further to keep laterality-and Laterality -going for a while. In the book, I argued that autonomous speech may have emerged as a cultural invention, just as writing did tens of thousands of years later. I now have reason to believe that it may have been the result of a genetic mutation. Over the past 15 years or so, there has been considerable interest in a large family, known as the KE family, with a genetic disorder of speech and language. The disorder is transmitted as an autosomal-dominant monogenic trait, encoded by a mutation on a gene on chromosome 7 known as FOXP2 (Lai, Fisher, Hurst, Vargha-Khadem, & Monaco, 2001). Some have argued that this gene is a grammar gene (e.g., Pinker, 1994), but although those affected have difficulties with both receptive and expressive grammar, the core deficit is more likely one of articulation (Watkins, Dronkers, & Vargha-Khadem, 2002). The FOXP2 gene underwent changes in hominins at some point subsequent to the split between hominin and chimpanzee lines, and probably within the past 100,000 years (Enard et al., 2002). Enard et al. write that their discovery "is compatible with a model in which the expansion of modern humans was driven by the appearance of a more-proficient spoken language" (p. 871). Evidence from the analysis of mitochondrial DNA suggests indeed that modern non-Africans have a common African ancestor dating from perhaps as recently as 52,000 years ago (Ingman, Kaessmann, Pääbo, & Gyllensten, 2000), implying that migrations from that time eventually replaced all other hominins outside Africa. Perhaps the mutation of the FOXP2 gene was the final adjustment that allowed speech to become autonomous, freeing the hands for the development of technologies that allowed our African forebears to populate the world-and wipe out all other hominins. And we've been doing the same sort of thing ever since. Just why the hominins of 50,000 years ago achieved such dominance has been regarded as something of a puzzle, especially since the evidence from mitochondrial DNA also suggests that our species originated in Africa much earlier, some 170,000 years ago (Ingman et al., 2000). Some have argued that it was the achievement of full-blown language that did the trick, but is it really likely that language could have evolved so late? Language itself seems so much part of being human that it is difficult to believe that Homo sapiens ever existed without it. My surmise, then, is that the critical development was the emergence of autonomous speech, whether as an invention (like writing) or as the result of a mutation that enhanced articulatory sequencing. True generative language itself, I suggest, evolved gradually over the past 2 million years, but it was the final emergence of autonomous speech that gave our forebears the edge. Why was autonomous speech so decisive? I attempted several answers to this question in the book, some of which may smack (sic) of special pleading. The most compelling advantage, I argued, was indeed the freeing of the hands for the enhancement of technology and the creation of a vocal discourse that could run in parallel with manual demonstration, leading to more sophisticated pedagogy. This argument may well need more work, but I nevertheless venture the more general suggestion that variations in the mode of communication may have been utterly critical to the evolution of our species. The later invention of writing also had a major impact, perhaps setting us back on a path towards visual communication. As we retreat further into the solitude of email and text messaging, we may be reduced again to a form of communication that even Kanzi could cope with. I am under no illusion that the gestural argument will be easy to sell, but take comfort in the fact that things are not always what they seem to be. Physicists tell us that matter consists mostly of the space between tiny particles, but it doesn't feel like that. One might even be tempted to argue, with Lewis Wolpert (1993), that science is most convincing precisely when it is unnatural, or counterintuitive. But it may have been a mistake to be rude about the Australians, some of whom are my best critics. I really have nothing against them, except that they keep winning at rugby, and don't even care much about the game. MICHAEL C. CORBALLIS University of Auckland, New Zealand