The reconstruction of the evolution of human spoken language

Mary LeCron Foster

The regularities of sound-change reconstructed by the comparative method provide the most reliable diagnoses of remote linguistic relations; but these are limited to 'families', or, in a few cases, 'stocks' made up of interrelated families. Broader groupings, 'phyla' or 'super-stocks', are suggested on the basis of typological relations, rather than on firmly established sound-correspondences. The basis for going even further and attempting to reconstruct a single prototype for all the world's spoken languages is not agreed upon; but the reconstruction should reflect systematic correspondences in sound and meaning throughout, whether insights were initially gained from typological studies of phonology and/or from internal reconstructions. Hypotheses must show system. While individual meanings underlying reconstructed forms need not be identical, differences should be minimized. Once correspondences are firmly established, culturally influenced semantic variations are useful in assessing degrees of interrelationship among languages.

Pursuing the monogenetic reconstruction through this bare-bones phonemic approach, refined by a series of simplifications, leads to the startling hypothesis that the sounds of which the VC and CVC roots are composed were originally themselves meaning-bearers. These phememes, as they are termed, were minimal units of sound whose meaning derived from the shaping and movement of the articulatory tract. In other words, the phonemes of language, as well as the combinations into which they unite within the word were originally not arbitrary signs, but abstract, highly motivated analogical symbols.