Strength Asymmetry and Positional Licensing: An Optimality Theoretic Account of Assamese Regressive Voicing Assimilation

Abstract

An attempt is made in this paper to examine whether strong or weak licensing capacity in a phonological domain is an inherent abstract property assigned by UG irrespective of languages or conditioned by phonetic factors. It is the normal case for languages to have homogeneous voice clusters, which are created by spreading both values of the [+/-voice] feature over the entire cluster, usually in a regressive fashion. Assamese exhibits the instance of regressive voicing assimilation, in which it is seen that the segment occurring in the coda position tends to agree in terms of feature [voice] with the following voiced obstruent in the onset position. But this regressive voicing assimilation is blocked by nasals and liquids. Since obstruent clusters agree in terms of sonority value they display assimilaton which is blocked among the segments of different class such as nasals, laterals and rhotics which are characterized by diverse sonority values. Voicing assimilation is functional among the segments of same sonority value and this is blocked between the segments of asymmetric sonority value. The data in this paper are collected from primary sources: ten native speakers of Assamese.