• Volume/Page
  • Keyword
  • DOI
  • Citation
  • Advanced
   
 
 
 

Journal of the Acoustical Society of America

SECTION LISTING

BROWSE VOLUMES

Year Range: 
Search Issue | RSS Feeds RSS
Previous Issue Next Issue

Nov 1975

Volume 58, Issue S1, pp. S2-S132

Return to All Sections
back to top
RSS Feeds
back to top Session EEE. Speech Communication XI: Linguistic Studies
Contributed Papers
FREE

Cross‐linguistic study of categorical perception for lexical tone (A)

Stephen W. Chan, Chiu‐Kuang Chuang, and William S.‐Y. Wang

J. Acoust. Soc. Am. Volume 58, Issue S1, pp. S119-S119 (1975); (1 page) | Cited 1 time

Online Publication Date: 11 Aug 2005

Full Text: | Download PDF

Show Abstract
This paper reports on tone perception by Chinese and American subjects and presents evidence that tones are perceived categorically by tone language speakers. Subjects were asked to identify and discriminate (ABX) synthetic speech stimuli which varied in acoustically equal steps through a range sufficient to produce the Mandarin “rising” and “level” tones. Chinese subjects showed a categorical boundary located near the middle of the stimulus space, i.e., there is a good correspondence between the crossover points in the identification functions and the peaks in the discrimination functions. American subjects exhibited a boundary closer to the level tone; this may be due to perception in a psychophysical mode. When tested on nonspeech stimuli having the same F0 characteristics, Chinese subjects showed a similar categorical boundary. Trained Chinese subjects having comparatively small JND for tones tend to perceive the stimuli only psychophysically. It would be of interest to determine if the acoustical distance between stimuli is sufficiently decreased for the discrimination task, and whether these subjects would also show a more prominent peak in their discrimination functions indicating categorical perception for tone. [Work supported by NSF, Grant no. SOC‐05798A01.]
FREE

Acoustic correlates of stress in Russian (A)

Edward T. Purcell and Nancy S. Brandt

J. Acoust. Soc. Am. Volume 58, Issue S1, pp. S119-S119 (1975); (1 page)

Online Publication Date: 11 Aug 2005

Full Text: | Download PDF

Show Abstract
Previous studies have examined the effects of various acoustic parameters on the perception of stress in English and other languages. Morton and Jessem [Lang. Speech 8, 159–181 (1965)] found that a raised F0 produced far greater effects on stress perception than did variations in intensity or duration. Fry [J. Acoust. Soc. Am. 27, 765–768 (1955)] showed that duration is more effective than intensity in cuing stress, though both are involved. Fry [Proc. 5th Int. Congr. Phon. Sci., Munster, 1964, 306–311 (1965)] found that vowel formant frequency is less effective than intensity in the determination of English stress. In Russian, vowel quality is directly and predictably related to stress placement, resulting in perceptual mergers of unstressed vowels. We synthesized pairs of Russian bisyllabic words with differing combinations of F0, duration, intensity, and vowel ferment frequency modifications one one syllable with the other syllable remaining untmodified. Five native speakers of Russian were asked to identify the randomly presented stimuli as fitting one of two possible grammatical frames. A nonsense‐word pair also was synthesized and similarly presented. Results are discussed in terms of the relative effects of the four parameters and combinations thereof on stress perception in Russian. [Work supported in part by ONR.]
FREE

Phonological rules influenced by boundaries (A)

Hector Javkin

J. Acoust. Soc. Am. Volume 58, Issue S1, pp. S119-S119 (1975); (1 page)

Online Publication Date: 11 Aug 2005

Full Text: | Download PDF

Show Abstract
Coker [J. Acoust. Soc. Am. 55, 476 (1974)] has argued that (1) the indestructibility of speech is due to the large measure of redundancy that speech contains and (2) that an important role of various phonological processes that create redundancy is to mark boundaries. A survey in 13 languages of rules that are conditioned by boundaries does not seem to support claim (2). Although 183 out of 524 rules examined made reference to boundaries, these did not create reliable boundary markers (i. e., features, segments or combinations of same that only and always occurred at boundaries). Furthermore, there were a number of cases in which a language had the opportunity to mark boundaries merely by blocking the application of a rule, but failed to do so. This suggests that phonological processes occurring at boundaries stem from the same phonetic constraints occurring elsewhere. A detailed phonetic examination of these languages might reveal the existence of reliable boundary markers, but the macroscopic view provided by phonelogical rules fails to do so. [This work supported in part by the National Science Foundation, and used the resources of the Stanford Phonology Archive.]
FREE

How can vowel systems differ? (A)

George Papçun

J. Acoust. Soc. Am. Volume 58, Issue S1, pp. S119-S120 (1975); (2 pages)

Online Publication Date: 11 Aug 2005

Full Text: | Download PDF

Show Abstract
On what parameters may vowel systems differ? To seek a partial answer to this question, formant data from two five vowel languages, Spanish and Japanese, are compared. Three mode factor analysis is demonstrated as a method of normalizing the data and extracting parameteric differences between languages and among speakers within languages.
FREE

Sentence prosody in isolated and in story contexts (A)

Lloyd H. Nakatani

J. Acoust. Soc. Am. Volume 58, Issue S1, pp. S120-S120 (1975); (1 page)

Online Publication Date: 11 Aug 2005

Full Text: | Download PDF

Show Abstract
Isolated sentences read from a list are often used as test material in speech experiments. But if isolated sentences differ noticeably from the same sentences read in a story, then the generality of observations on isolated sentences is severely restricted, and the construction of test material which permit broader generality is troublesome. To see if isolated and story sentences were perceptually distinguishable, listeners were presented with stories read normally and stories spliced together from sentences read from a randomized list. The listeners had difficulty identifying the stories as normal or spliced. This and other results affirm that speech features observed in isolated sentences can be reasonably generalized to sentences in a coherent context.
FREE

Effect of normalization and transformation in the processing of F0 for analyzing American intonation (A)

Yukio Takefuta

J. Acoust. Soc. Am. Volume 58, Issue S1, pp. S120-S120 (1975); (1 page)

Online Publication Date: 11 Aug 2005

Full Text: | Download PDF

Show Abstract
As any researchers who studied intonation know, this speech signal is an extremely difficult signal to analyze scientifically. Everytime they study it they find the necessity of studying more basic facts about it. Having learned this fact, this investigator collected samples of the simplest intonation, those of one word sentences, and examined the effect of normalization and transformation in the digital processing of physically extracted data F0 for the valid analysis and accurate detection of interrogative signals in American intonation. When optimal methods of normalization and transformation (ten methods tested) were applied, all interrogative intonations of ten different sentences (produced by 60 American speakers and validated by five native speakers of English) were perfectly distinguished from noninterrogative intonations. After these methods of normalization and transformation were applied to the physically extracted data, the basic interrogative intonation pattern and its variant for each of the ten one‐word sentences were obtained.
FREE

Dialect variation in pitch contours on stressed syllables (A)

M. H. O'Malley and M. Yaeger

J. Acoust. Soc. Am. Volume 58, Issue S1, pp. S120-S120 (1975); (1 page)

Online Publication Date: 11 Aug 2005

Full Text: | Download PDF

Show Abstract
The pitch contour is often considered to be an important signal, if not the principal acoustic signal, of linguistic stress. While it has been recognized that pitch dips as well as pitch rises can indicate stressed syllables, stress detectors for automatic speech recognition systems have often been based on the detection of pitch rises. The present experiment measured pitch contours on stressed syllables for speakers of several dialects. Both formal and informal styles of speech were used. It was found that the use of pitch contours to signal stress is a function of dialect. In particular, there exist dialects for which duration or amplitude rather than pitch is usually the cue for stressed syllables. The results of this experiment are further evidence that dialect variation and speech style must be considered in studies of the acoustics of speech.
FREE

Modifying fundamental frequency envelopes (A)

A. K. Melby, R. Millett, W. J. Strong, and E. G. Lytle

J. Acoust. Soc. Am. Volume 58, Issue S1, pp. S120-S120 (1975); (1 page)

Online Publication Date: 11 Aug 2005

Full Text: | Download PDF

Show Abstract
An experiment was designed to investigate the hypothesis that changes in fundamental frequency alone are sufficient to cue several readings of a sentence. A sentence was recorded and digitized in four versions and analyzed by a predictor coefficient method. First, the sentence was synthesized with its natural F0 envelope. Then it was synthesized with its F0 envelope being replaced by artificial ones patterned after the envelope of one of the three stressed versions. In each case, the synthesized output was clearly perceived to be the stressed version whose F0 envelope had been borrowed, although other prosodic features such as duration, pause, and amplitude were unaltered. This experiment seemed to support the importance placed on F0 by a recent article by Olive [J.P. Olive, J. Acoust. Soc. Am. 57, 476–482 (1975)].
Return to All Sections
Close

Home Publications Meetings Contact ASA Site Map Back to Top

Copyright © Acoustical Society of America

close