Acrobat Reader 3.0 or
higher to read .pdf files.You need
Acrobat Reader 3.0 or
higher to read .pdf files.
ABSTRACT
An acoustic analysis, synthesis, and perception experiment were performed to investigate the parameters that determine the aspirated/tense distinction in Korean alveolar fricatives. Korean has two types of alveolar fricatives, both of which are voiceless. One is called an aspirated fricative /s/, and the other a tense fricative /s?. Five sets of minimal pairs with five Korean vowels /a, e, i, o, u/ were chosen for the investigation.
Analysis results showed that among the various parameters that were measured, only the duration of the aspiration segment was consistent across all speakers in differentiating the fricatives in mid and low vowel contexts. The aspiration of tense fricatives was less than 21 miliseconds. Other parameters showed some trends, but they were not as consistent as the duration of the aspiration segment. When the aspiration duration was not the major cue for the distinction, it seems that other parameters play a role.
An aspirated alveolar fricative /s/ followed by a vowel /a/ was synthesized using the Klatt Synthesizer based on measurements taken from the speakers. When the aspiration between the fricative and the vowel was removed, perception shifted from the aspirated category to the tense.
In the perception experiment with natural utterances, the word /sada/ from eight speakers was manipulated by removing the aspiration segment in 10msec increments from the center of the aspiration. The tokens were presented to twenty native speakers of Korean for identification tests. The results showed that for most listeners, the duration of the aspiration segment alone can act as the primary cue for the aspirated/tense distinction.
Adobe acrobat files for downloading
Hyperlinked table of contents (0.97M)
Chapter 1 (221k)
Chapter 2 (1.01M)
Chapter 3 (544k)
Chapter 4 (225k)
Chapter 5 (178k)
References & Appendices (301k)