ACOUSTIC CHARACTERISTICS OF MESKWAKI STOPS
University of Iowa
Thursday, December 3, 2015
The Algonquian language Meskwaki is an endangered language spoken fluently by about 200 older members of the Meskwaki tribe, which has communities situated in Iowa, Oklahoma, and Kansas. To date, there has not been any formal phonetic analysis of the Meskwaki language. A definitive descriptive overview of the language was written by William Jones in 1911, and further descriptive work on the phonology of the language was conducted by Ives Goddard in 1991 and in the early 2000s. Jones (1911) describes Meskwaki as having both a voiced and voiceless stop series, as well as a pre-aspirated series of stops. Goddard (1991), however, includes only voiceless stops, no voiced stops, and claims that the pre-aspiration in Meskwaki is non-phonemic and highly dependent on individual variation. Later research argued that Meskwaki, and Algonquian languages in general, do not have true pre-aspiration, but rather hC clusters (Helgason 2002, Silverman 2003).
However, a pilot study conducted in Meskwaki as spoken at the Settlement in Tama, Iowa determined that while Meskwaki does not appear to have voiced stops, at least not in initial position, there is a regular distribution pattern of pre-aspirated and unaspirated voiceless stops. The goal of this research was to build on the results found in the pilot study to determine the distribution of pre-aspirated and unaspirated voiceless stops in Meskwaki, and provide an instrumental accoustic analysis of the pre-aspirated stops in Meskwaki to determine whether the language possesses true pre-aspirated stops, or hC clusters. To this end, stops were examined in initial and medial position in isolated words, as well as in connected speech.
In answer to the second research goal, this study found that the pre-aspirated stops in Meskwaki are produced with the characteristic limited distribution, glottalic friction before the stop closure, and sonorant devoicing that Helgason (2002) used as the guidelines for pre-aspiration. This pre-aspiration was also consistent across the speakers used for the research. By the standard set out by previous research, Meskwaki has true pre-aspiration, and not hC clusters. In answer to the first research goal, this study found that in utterance- and word-initial positions, there is indeed no clear contrast in Meskwaki stops. All initial stops are produced with short-lag VOT durations. However, in word-medial contexts, there is a clear distinction between stops that are produced with short-lag VOT and stops produced with pre-aspiration. Furthermore, data from connected speech showed that some words that were produced with short-lag VOT in utterance-initial contexts were pre-aspirated in utterance-medial contexts.
Based on the data collected and analyzed in this research, this paper proposes that the Meskwaki language has both true pre-aspirated stops and short-lag stops. Additionally, it is suggested that in Meskwaki, the feature of contrast may be [spread glottis]; specified stops may be realized with pre-aspiration in medial contexts and short-lag due to phonetic constraints in initial contexts, while unspecified stops are realized as short-lag throughout.
Please join us at 3:30 in 571 EPB for refreshments.