Defining initial strength in clusterless languages in Strict CV

Abstract

Strength is defined by position in Strict CV so that the strength associated to the initial position and the weakness associated to ‘codas’ is a result of their position with respect to empty positions in a configuration of strictly alternating C and V positions. 1 Within this configuration, government and licensing relations determine which positions are regarded as strong or weak. With respect to strength in initial position this must interact with a postulated initial (melodically empty) CV unit (Lowenstamm 1999) that marks the beginning of a word and that facilitates the categorisation of languages into those with only sonority-increasing clusters and those with either sonority-increasing or decreasing clusters in initial position. 2 While these findings provide profound insight into how grammar characterises languages into two types with respect to the initial cluster type attested and makes predictions on the expected strength in initial position in these languages, it remains to be seen whether languages without clusters can draw on the mechanisms developed to define strength in initial position. This paper investigates whether this expanded theory sheds light on strength in initial position in languages without clusters, here dubbed clusterless languages. It will be shown that, contrary to what is seen in languages with clusters where the absence of the initial CV unit implies a weak initial position, clusterless languages provide evidence that although the initial CV unit is absent, the initial position still exhibits properties of strength. This lack of predicted weakness will be accounted for by parameterising proper government and suggesting that this governing relation has no role to play in clusterless languages. In the absence of proper government, an