One very different piece of evidence was unveiled in 2002 (this was mentioned earlier, in a different context). A team led by Svante Paabo at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany, announced in August that year that it had identified two critical mutations which appeared approximately 200,000 years ago in a gene linked to language, and then swept through the population at roughly the same time anatomically modern humans spread out and began to dominate the planet. This change may thus have played a central role in the development of modern humans’ ability to speak.
33 The mutant gene, said the Leipzig researchers, conferred on early humans a finer degree of control over the muscles of the face, mouth and throat, ‘possibly giving those ancestors a rich new palette of sounds that could serve as the foundation of language’. The researchers did not know exactly what role the gene, known as FOXP2, plays in the body, but all mammals have versions, suggesting it serves one or more crucial functions, possibly in foetal development.34 In a paper published inEven more controversial than the debate over when language began have been the attempts to recreate early languages. At first sight, this is an extraordinary idea (how can words survive in the archaeological record before writing?) and many linguists agree. However, this has not deterred other colleagues from pushing ahead, with results that, whatever their scientific status, make riveting reading.
One view is that language emerged in the click sounds of certain tribes in southern Africa (the San, for example, or the Hadzabe), clicks being used because they enabled the hunters to exchange information without frightening away their prey on the open savannah. Another view is that language emerged 300,000–400,000 years ago, and even 1.75 million years ago, when early man would sing or hum in a rhythmical way. Initially, these sounds were ‘distance calls’, by which males from one group attracted females from another group (as happens with some species of chimpanzee), but then the rhythmic chanting acted as a form of social bonding, to distinguish one tribe from another.
From such other anthropological evidence as exists, from contemporary hunter-gatherer tribes, we find that there is about one language for every thousand or two thousand people (there were around 270 Aboriginal languages in Australia when that continent was discovered by Europeans).
35 This means that, at the time man crossed from Siberia to Alaska, when the world population was roughly 10 million,36 there may have been as many languages in existence then as there are today, which is – according to William Sutherland, of the University of East Anglia – 6,809.37 Despite this seeming handicap, some linguists think that it is possible to work back from the similarities between languages of today to create – with a knowledge of pre-history – what the original languages sounded like. The most striking attempt is the work of the American Joseph Greenberg who distinguishes within the many native American languages just three basic groupings, known as Eskimo-Aleut, Na-Dene and Amerind. His investigations are particularly noteworthy when put alongside the evidence, mentioned earlier, that there were three migrations into the Americas from Asia.38 2 The latest DNA evidence, however, suggests there were not three but five waves of migration from Siberia into America, one of which may have been along the coast.40 This evidence suggests that the first Americans may have entered as early as 25,000 years ago – i.e., before the Ice Age, and meaning that these pioneers