If you were told to find out first-hand about the roots of a plant, you would be likely to go out into the garden, stand at the level of today’s topsoil, and dig downwards. In this chapter, we tackle the apparently common-sense idea that to find out about the roots, or ultimate origins of languages, we should equally start from languages we know today and work backwards to their roots; and we show that this approach cannot work. First, the good methods we have for linguistic reconstruction (notably the comparative method) appear to be time-limited and cannot take us back far enough to be of evolutionary relevance. There are less-constrained methods, but we cannot rely on their results. Second, and perhaps more important, any such method will necessarily be working at the level of behavioural, output language structures, and we have seen that if we really want to talk about evolution, we need to get behind such superficial behavioural characteristics to the underlying physical and neurological structures, and ultimately to the genetic instructions which have configured them. We turn, therefore, from linguistic reconstruction to the comparative method in biology, and consider some similarities and differences between human language and aspects of other primate communication systems.
Going backwards to move forwards
When Jespersen (1922; discussed in 1.3.2 above) has finished noting his disappointment with early ‘speculative’ theories of language origins, of the bow-wow and yo-he-ho variety, he suggests that we should instead turn to evidence from the histories of languages. On the basis of earlier stages of present-day languages, and reconstructions of languages which no longer exist, we may ‘attempt … step by step to trace the backward path. Perhaps in this way we may reach the very first beginnings of speech’ (1922: 418). Jespersen suggests, as we saw earlier, that change typically involves reductions in the phonological shapes of words, and in word length, along with a trend towards greater analysis, and division into small, individual forms, each with its own meaning. He therefore proposes that ‘Primitive linguistic units must have been much more complicated in point of meaning, as well as much longer in point of sound, than those with which we are most familiar’ (1922: 425); moreover, these long words were rather unstructured internally, with each expressing the meaning of a whole sentence today.