出版社：Department of Linguistics, University of Toronto
摘要：Hamann (2003) identifies several articulatory properties of retroflexion, three of which correspond to distinctive features commonly employed in the literature: posteriority ([–anterior]), apicality ([–distributed]) and retraction ([+back]). I use phonological activity as a diagnostic to determine which of these features are phonologically distinctive for retroflex segments in Indo-Aryan languages. The evidence suggests that retroflex segments in these languages are distinctively apical ([–distributed]), and potentially retracted ([+back]) at a post-lexical level. There is no evidence for [–anterior] at any level of representation. In light of similar findings reported in Dravidian and Australian languages with larger coronal place inventories, I argue that this is not a case of contrastive (under) specification. Rather, I suggest that posteriority ([–anterior]) may be a universally redundant and non-essential property of retroflexion that is derived from the combination of apicality ([–distributed]) and retraction ([+back]). Some implications for models of coronal place features are briefly discussed.
