Abstrakt
Classification of related languages as to the level of mutual genetic proximity relies on a number of criteria, the principal one being the character and number of shared innovations. As singular innovations are likely to emerge in related languages independently, the risk of misinterpreting homology for homoplasy is high.
In this respect, suppletion, as the extreme of morphological irregularity and therefore an innovation least likely to emerge independently in the semantically and etymologically identical paradigms, may help to support or invalidate hypotheses. In this article, I examine a number of shared suppletive paradigms to show how their distribution in the family may shed light on its bifurcation.