Publication date: September 2016
Source:Journal of Phonetics, Volume 58
Author(s): Mary Stevens, Jonathan Harrington
In contemporary spoken English, /s/ can resemble a post-alveolar fricative when it occurs in /str/ clusters e.g. street. /s/-retraction in /str/ is known to be widespread in North American English, but the question of how this sound change comes about has attracted only a very small amount of empirical research. This paper investigates the phonetic pre-conditions for /s/-retraction based on the results of two experiments conducted with Australian English. Study 1 shows that the first spectral moment (M1) for /s/ is lower in /spr, str, skr/ and slightly lower in /sp, st, sk/ than in pre-vocalic position. Temporal variation in first spectral moment trajectories suggests that the articulatory movements associated with the rhotic are timed relatively earlier in /str/ vs. /spr, skr/ clusters. Study 2 tested native listener categorization of sibilants produced by multiple talkers. Results show that sibilants originally produced in stream and steam elicit /ʃ/ responses when spliced into pre-vocalic contexts. There was an interaction with talker gender, with male voices eliciting more /ʃ/ responses. Thus, the conditions for sound change to occur by which /s/ becomes /ʃ/ originate in a synchronic bias for /s/ produced in /sC(r)/ clusters to show increased auditory similarity to /ʃ/. Australian English shows the pre-conditions for /s/-retraction but is not currently undergoing this sound change.
from Speech via a.lsfakia on Inoreader http://ift.tt/2bkD7o2
via IFTTT
Δεν υπάρχουν σχόλια:
Δημοσίευση σχολίου