LLTI Archives

April 2001, Week 4

LLTI@LISTSERV.DARTMOUTH.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
LLTI-Editor <[log in to unmask]>
Reply To:
Language Learning and Technology International Information Forum <[log in to unmask]>
Date:
Thu, 26 Apr 2001 08:26:44 EDT
Content-Type:
text/plain
Parts/Attachments:
text/plain (40 lines)
--- Forwarded Message from Judy Shoaf <[log in to unmask]> ---

>Date: Wed, 25 Apr 2001 09:24:53 -0400
>To: Language Learning and Technology International Information Forum               <[log in to unmask]>
>From: Judy Shoaf <[log in to unmask]>
>Subject: Re: #6078.3 Voice analysis / waveform (!)

------------------
Derek Roff's point about the problems of matching waveforms is a good one.
However, I have heard CAN-8 enthusiasts say that trying to match the
waveforms in that software keeps the students working at pronunciation. When
I have played around with it I have noticed that what I try to match is the
duration of each sound, which is a bit funny if you consider that the models
are speaking very slowly. But visual feedback is I think a good thing and
encouraging to students.

Here at the University of Florida we have had a few instructors (in
Linguistics) who really understand the different types of images that can be
generated by the spoken word, and one instructor even worked one-on-one with
students to correct French pronunciation using computer speech analysis
software. But this involved being able to pick the image that would target
the problem to be addressed and showing the student the bit of the image
that needed to be imitated (see that high point there, followed by a dip?
that's the "r" in this word..). 

There is a program for ESL called TEAM (technologially enhanced accent
modification), of which we have a beta version here. It's valuable because
someone already did the work of targeting the images to the specific
pronunciation problems. If you want to work on intonation, you get one kind
of pattern to match that reflects intonation; if you are working on vowels,
you get a different image which shows where in your oral cavity you are
pronouncing that vowel, more or less. 

I can imagine a nice little CD-ROM for any given language that would run
through the basics of pronunciation in this way, for relatively advanced
students aware of their deficiencies and targets. 

Judy Shoaf
University of Florida, Gainesville

ATOM RSS1 RSS2