LINGUIST List 9.1092

Sun Aug 2 1998

Sum: Compressed Speech and Phonological Features

Editor for this issue: Martin Jacobsen <martylinguistlist.org>


Directory

  1. Susan Fischer, Compressed Speech and Phonological Features

Message 1: Compressed Speech and Phonological Features

Date: Mon, 27 Jul 1998 18:03:00 +0000
From: Susan Fischer <sdfncrritvax.isc.rit.edu>
Subject: Compressed Speech and Phonological Features

I posted the following query last week:


Date: Fri, 17 Jul 1998 07:30:29 +0000
>From: Susan Fischer <sdfncrritvax.isc.rit.edu>
Subject: Phonological Processing

Does anyone know of literature that addresses the following question?

What phonological or phonetic features are more or less difficult to
perceive as the speed/compression of speech increases? I'll be glad
to summarize any responses for the list. TIA


Thanks to the following for their responses, which are given in edited
form below:
		Martha McGinnis <marthajoMIT.EDU>
		Natasha Warner <Natasha.Warnermpi.nl>
		Dan Faulkner <Dan.Faulkneraculab.com>
		"James L. Fidelholtz" <jfidelsiu.buap.mx>
		Esther Janse <Esther.Janselet.uu.nl>


 
As you may know, Paula Tallal of Rutgers University has done research
on phonological processing at varying speeds. (Her theory is that
dyslexia arises from slowness of phonological processing -- a theory
she's successfully marketed to parents of dyslexic kids). I'm not
sure whether she's looked in detail at different phonological
features, but you could check out her work. Some references are given
on her homepage:

http://www.cmbn.rutgers.edu/cmbn/faculty/tallal.html

Best wishes,
Martha McGinnis

________________
 

The article by Furui in JASA (1986) might be useful. It's not about
fast speech, but in the latter part of the article, he investigates
how much of the signal listeners need in order to perceive various CV
sequences. The reference is:

Furui, Sadaoki. 1986. On the role of spectral transition for speech
perception. Journal of the Acoustical Society of America 80:1016-1025.

Natasha Warner
______________


I do not know of any work that specifically relates to features
becoming more or less difficult to perceive with changes in speech
rate, but there is a lot of evidence for prosodic phonetic features
changing with speech rate. One paper that relates speech rate to the
segmental string is:

O'Dell, M. 'Speech Tempo and quantity perception' from 'Papers in
research' ed. Pertti Hume. Jyvaskyla (1985).

Some that relate to phonetic prosodic effects are:

Rietveld & Gussenhoven, 'Perceived speech rate and intonation.'
(unfortunately unpublished I think, but I might be able to get a copy
for you.'

van Santen and Hirshberg 'Segmental effects on the timing and height
of pitch contours.' from ICSLP 1994, pp 719-722.

Steele,S.A. 'Nuclear accent F0 peak location: effects of rate, vowel,
and number of following syllables' from JASA (1986), suppleme.t 1, 80

Caspers, J 'Pitch movements under time pressure'. HIL (1994)

If these are useful, I can send you some more. It depends on whether
or not you count prosody as a phonological domain I suppose - I do,
but sometimes people prefer to keep it to segments, don't they?

Dan Faulkner 

Computational Spoken Language Scientist
Aculab plc
Tel: +44 1908 273 933
Fax: +44 1908 273 801
www: http://www.aculab.com

_______________


 This may not go quite to the question you are asking, and I don't
have any specific references for you at the moment, but one thing that
has been known for at least 30 years is that vowels can be compressed
with less loss than consonants (in general), and eg /s/ and other
fricatives lose less than stops. In the late 60s, I actually used a
program which took these factors into account. If I recall correctly,
if you do sensible compression, you can cut out about 50% of the
signal and still understand the result. You lose quite a bit if you
just lop off, say, every other millisecond of the signal, however. I
think the program was developed in the DOD unit I was working in at
the time. I haven't kept up with this, but there must be some
programs around to do this. Sorry not to be more explicit, but I hope
this helps some.
 Jim

James L. Fidelholtz e-mail: jfidelcen.buap.mx
Maestri'a en Ciencias del Lenguaje
Instituto de Ciencias Sociales y Humanidades
Beneme'rita Universidad Auto'noma de Puebla, ME'XICO

________________

I read your call on the Linguist List in which you asked for
literature references on the perception of phonological or phonetic
features in time-compressed speech. Actually, I am planning to do a
perception study on this subject myself and I have also been trying to
find relevant literature. So far, I have not found much but I do have
the following references for you, although they are mainly on the
production and spectral aspects of fast speech.

Dani Byrd (1996), 'Influences on articulatory timing in consonant
sequences', Journal of Phonetics 24.

Byrd and Tan (1996), 'Saying consonant clusters quickly', Journal of
Phonetics 24.

Zsiga (1994), 'Acoustic evidence for gestural overlap in consonant
sequences', Journal of Phonetics 22, 121-140.

Miller (1981), 'Effects of speaking rate on segmental distinctions',
in 'Perspectives on the study of speech' by Peter D. Eimas and Joanne
L. Miller (eds.), Hillsdale, New Jersey.


With regard to the perception of fast speech, there have been numerous
studies on the perception of Voice Onset Time continua and if you are
interested in those, I would have some more references for you.

Esther Janse
Esther.Janselet.uu.nl
Utrecht Institute of Linguistics OTS
Utrecht
The Netherlands

I am not sure whether these studies on the perception of
voiced/voiceless distinctions may be relevant to your analyses, so I'd
better give you some any way.

Miller & Volaitis (1989), Perception and Psychophysics 46 (6), 'Effect
of speaking rate on the perceptual structure of a phonetic category'.

Kidd (1989), Journal of Experimental Psychology: Human Perception and
Performance 15 (4), 'Articulatory-rate context effects in phoneme
identification'.

Miller & Grosjean (1981), Journal of Exp. Psychology: Human Perception
and Performance 7 (1), 'How the components of speaking rate influence
perception of phonetic segments'.

There are also people in France (Jacques Mehler, Emmanuel Dupoux and
Christophe Pallier) working on adaptation to time-compressed
speech. Some of their articles can be downloaded from the internet
from the addresses listed below.

http://www.ehess.fr/centres/lscp/persons/dupoux.htm#listpub

http://cogito.lscp.ehess.fr/web/persons/pallier/articles.htm

Esther Janse

___________________

Thanks again to all who responded. My colleagues and I have completed
a study in which we examined time-compressed signing and were
interested in comparisons with speech, especially in error analysis

Susan Fischer - Susan Fischer 
Dept. Of Applied Language & Cognition Research
52 Lomb Memorial Drive 
Rochester, NY 14623-5604
e-mail: sdfncrrit.edu
NTID/RIT phone: 1-716-475-6558 (v/TTY)
fax: 1-716-475-6500
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue