The nature of the schwa/zero alternation in French clitics: Experimental and non-experimental evidence

This article examines the phonological status of schwa in clitics, in particular whether or not schwa should be included in their lexical representation. Several distributional and experimental arguments pointing to the lexical status of clitic schwas are reviewed and are shown to be inconclusive, due to the existence of additional data that suggest a different interpretation not involving underlying schwas. The discussion includes experimental results that fail to show residual lip rounding in the vicinity of an omitted schwa at clitic boundaries, contra Barnes and Kavitskaya's (2002) previous claim. In the absence of evidence to the contrary, the non-contrastive nature of clitic schwas militates against their underlying status.