Two types of [NT]s in Panãra: Evidence for temporally ordered subsegmental units

Abstract

This paper discusses articulatory and perceptual phonetic data on Panãra (ISO code: kre), a Jê language of Central Brazil, supporting the existence of a previously undocumented phonological distinction between two types of [NT] segments. Panãra exhibits a distinction between partially nasalized consonants arising from two distinct phonological processes: post-oralization of nasal consonants (/m, n, ɲ, ŋ/ → [mp, nt, ns, ŋk]), and pre-nasalization of oral obstruents (/p, t, s, k/ → [mp, nt, ns, ŋk]). These [NT]s contrast in surface sequences of the type [ṼNTV], as in the minimal pair /mĩ-ŋɾɛ/ → [mĩŋkɾɛ] ‘caiman egg’ vs. /mĩ-kɾɛ/ → [mĩŋkɾɛ] ‘caiman burrow.’ This novel data provides clear evidence that phonological grammars can and do manipulate subsegmental units. The data is analyzed within the framework of Q Theory, a model of representational phonology which decomposes the segment [Q] into a series of three quantized, temporally ordered subsegments (q1 q2 q3) (e.g., Shih & Inkelas 2019). The tripartite architecture of Q Theory provides the level of granularity necessary to distinguish between post-oralized nasals and pre-nasalized stops, where the former are represented with two nasal subsegments followed by one oral subsegment (m1 m2 p3), and the latter are represented with a single nasal subsegment followed by two oral subsegments (m1 p2 p3). It is shown that, to model the distribution between the two [NT]s in Panãra, the grammar must crucially make use of constraints that reference subsegmental units.