LINGUIST List 14.1251

Sun May 4 2003

Sum: Interword vs. Intraword Coarticulation

Editor for this issue: Steve Moran <stevelinguistlist.org>

Directory

1. Charles Hoequist, Jr., Sum: interword vs. intraword coarticulation

This is a summary of reponses to a query I posted on March 28th
(Linguist 14.927) regarding coarticulation within and across word
boundaries.

My thanks to the following people for responding:

Joaquim Brandao de Carvalho
Cassie Mayo
Doug Whalen 
Sidney Wood
Robert Hagiwara
Alex Monaghan

First of all, my apologies for the delay; I have ordered some of the
recommended articles via interlibrary loan and can only conclude that
someone, somewhere, is copying them in longhand for me :). So I won't
be able to give a complete summary, but I can speak to what I have.

The question originated as a remark in a paper under preparation on
text-to-speech synthesis, saying roughly, "as is well known,
coarticulation is reduced across a word boundary". Given that most
TTS (all commercial TTS, as far as I know) now uses stored recorded
units, the question of differing coarticulation strongly influences
the inventory of stored units. However, when I tried to find evidence
for this offhand remark, I came up short. Hence the question.

Joaquim Brandao de Carvalho recommended Mark Aronoff's & Mary-Louise
Kean's book Juncture. We'll see if interlibrary loan can find it.

Cassie Mayo recommended a list of articles by Susan Nittrouer and
various co-authors, which are still working their way through
interlibrary loan:

(1989) "The Emergence of Phonetic Segments: Evidence from the Spectral
Structure of Fricative--Vowel Syllables Spoken by Children and
Adults", Journal of Speech and Hearing Research, 32 (120-132).

(1993) "The Emergence of Mature Gestural Patterns Is Not
Uniform---Evidence From an Acoustic Study", Journal of Speech and
Hearing Research, 36 (956-972).

(1995) "Children learn separate aspects of speech production at
different rates: Evidence from spectral moments", Journal of the
Acoustical Society of America, 97 (520-530). I had better luck finding
this, and though it doesn't directly address boundary issues in
coarticulation, I found it interesting for the evidence that
children's (co)articulation is not as precisely controlled as adults',
even for children up to 7 years.

(1996) "How Children Learn to Organize Their Speech Gestures: Further
Evidence From Fricative-Vowel Syllables", Journal of Speech and
Hearing Research, 39 (379-389).

(1992) "Age--related differences in perceptual effects of formant
transitions within syllables and across syllable boundaries", Journal
of Phonetics, 20 (351-382).

Sidney Wood directed me to his website with some nice x-ray-derived
animations of tongue coarticulation
(http://www.ling.lu.se/persons/Sidney) and publications based on the
same x-ray data:

Assimilation or coarticulation? Evidence from the coordination of
tongue gestures for the palatalization of Bulgarian alveolar stops.
In D. Recasens (ed), Lingual Data and Modeling in Speech Production
(Papers from the ACCOR Workshop on Lingual Data and Modeling in Speech
Production, Barcelona, December 1993). Journal of Phonetics 24,
139-164. 1996.

A cinefluorographic study of the temporal organization of articulator
gestures: Examples from Greenlandic. Paper presented to the First ESCA
Workshop on Speech Modeling and Fourth Speech Production Seminar,
Grenoble 1996. Published in P. Perrier, R. Laboissi�re, C. Abry and
S. Maeda (eds), Speech Production: Models and Data (Papers from the
First ESCA Workshop on Speech Modeling and Fourth Speech Production
Seminar, Grenoble 1996). Speech Communication 22, 207-225. 1997.

Robert Haqiwara directed me to Dr. Dani Byrd at UCLA, who turned out
to be a lot faster and more helpful than interlibrary loan. She was
kind enough to send me several papers and direct me to another, all
articulatory studies. Anyone who is interested should check her page
at http://www-rcf.usc.edu/~dbyrd/dbyrd.html

Now, to the question: to word boundaries make a difference in
coarticulation?

(pause)

You didn't think there was going to be a clear answer, now did you? Of
course there isn't. A major difficulty in investigating boundary
effects is that we have way too many possible explanations for any
phenomena we see. For example, Dr. Byrd's work contains strong
evidence that there are differences in the degree of temporal
articulatory control depending on whether or not the test utterance
spans a word boundary. But the data can be equally well or better
accounted for by a prosodic boundary, together with position in the
syllable (onset vs. coda).

Thanks to all who responded. This has opened up an interesting can of
worms.


Charles Hoequist, Jr.
Center for Communication Technology
Aalborg University