Although there is no master plan driving the process, it is clear that language change is a universal phenomenon, and patterned rather than random: certain kinds of changes recur in widely separated languages. Deutscher seeks to explain the underlying principles at work here, drawing on evidence from both real and reconstructed “proto” languages. He also seeks to show how those principles could account for a much earlier development, one linguists can only speculate about, since if it happened it took place so far back in prehistory as to be beyond reconstruction – the formation of human languages as we know them from the simpler systems that were their hypothetical precursors.
Deutscher imagines what he dubs a “Me Tarzan” stage of linguistic evolution, when humans communicated using a small number of words and some basic rules for ordering them, and applies what we know about language change to explain how such a “primitive” system might have acquired the complexity that is evident in even the oldest languages known to scholarship.
More here.