Hi,
here's the reference on the Ohala paper:
Ohala, J. J. 1994. Acoustic study of clear speech: a test of the
contrastive hypothesis. Proceedings, International Symposium on Prosody,
September 18, 1994, Pacific Convention Plaza Yokohama. 75-89.
It's available at:
http://trill.linguistics.berkeley.edu/users/ohala/index3.html
As for the ESL examples, even non-native English-speaking graduate
linguistics students in my phonetics class were amazed to find out that
the major difference between "can" and "can't" is usually reduction (in
the form of vowel quality and duration) rather than presence or absence of
an alveolar stop.
I also don't know of anything actually testing processing of speech in
which every word is hyperarticulated. It might be hard to test processing
difficulty independent of the overall slower speech rate such speech would
have. It seems like people working on speech synthesis might have tested
this at some point, since one could readily have a synthesizer produce
fully hyperarticulated speech and then speed it up. If hyperarticulated
speech were more intelligible than natural speech, that would be a
convenient application of synthesis.
Thanks,
Natasha Warner
*******************************************************************************
Natasha Warner
Assistant Professor, Department of Linguistics
University of Arizona
PO Box 210028
Tucson, AZ 85721-0028
|