Research

Ohio State Dissertations in Linguistics (OSDL)

Benjamin Xiaoping Ao (1993)

Phonetics and Phonology of Nantong Chinese

Advisor: David Odden

Download .pdf file [PDF]

Abstract:

This dissertation studies the phonological system of the Nantong dialect of Chinese and its implications for theoretical phonology.

The underlying sound system of Nantong Chinese consists of 38 consonants, 15 vowels and 5 tones. Instrumentally obtained phonetic descriptions are provided for post-alveolar and alveopalatal consonants, for cross-linguistically rare fricative vowels, and for tones.

To analyze segmental alternations and phonotatic constraints, the author proposes a new model of place feature geometry, which uses a unified set of place features for vowels and consonants. The features [labial] and [coronal] represent primary articulations when directly linked to the C-place node, or secondary articulations when linked through the V-place node. The feature [dorsal] dominates [coronal], and is always linked to the C-place node through the V-place node. This model can handle consonant-vowel interactions without being overgenerative.

To handle tonal alternations, a tonal feature and feature geometry model is also proposed. With one tone feature and two register features, this model can represent five and only five tone levels, which is the attested maximum of contrastive tone levels in any language. This model allows many tone sandhi processes to be analyzed as feature spreading or deletion, instead of arbitrary feature changing.

Based mainly on these models of features and feature geometry, a total of 13 segmental and 15 tonal rules are proposed to account for the morphophonemic alternations in this language. It is shown that each of these rules applies within a prosodically defined domain, i.e. the metrical foot, the phonological word, the phonological phrase, or the intonational phrase. It is also shown that the target of these rules may vary in different speech rates, suggesting that the metrical feet and the phonological words which define the domains of these rules differ in size and internal structure according to speech rate.