LINGUIST List 6.752

Thu 01 Jun 1995

Qs: Lawyers, Open vs Closed, Grad Students, Speech synthesis

Editor for this issue: <>

Directory

1. Michel Toporkoff, Lawyers and linguistics ?
2. Dan I. Slobin, Open vs. Closed Class
3. rosa graciela montes, Query: Graduate Students (linguistics) Organizations
4. Timothy Miller, Speech synthesis

Content-Length: 1006

Dear readers of "linguist list", I'm a french lawyer seriusly interested in
the legal aspects of linguistics.

Is there anybody else on this list sharing the same interest ?

Thank you

Michel

--
Michel Toporkoff, 6 rue J B Potin (92170) Vanves, France

Content-Length: 1079

Does anyone know who introduced the distinction between
"open" and "closed class" morphemes? I'd appreciate
references to the history and current status of this
distinction. Please respond to me directly and I'll post
a summary.

Dan Slobin (slobincogsci)
Dept of Psychology
University of California
Berkeley CA 94720-1650

Content-Length: 1435

On behalf of the students in the graduate program of Linguistics at the
University of Puebla (MEXICO) I would like information about linguistics
students organizations, associations and societies: forms of
organization, membership [local, regional, national], activities
sponsored [publications, meetings], links to similar professional
organizations, etc. In general, anything that might be useful to students
wanting to set up a similar type of organization/association.

Any help, pointers or comments would be appreciated

Dra. Rosa Graciela Montes
Director Maestria en Ciencias del Lenguaje
ICSyH
Univ. Aut. de Puebla (MEXICO)

rmontessiu.cen.buap.mx
rmontesrico.pue.udlap.mx

Content-Length: 2617

Hello, fellow linguists,

I'm working on a project in a microprocessor-based design class, and the
project that my team has decided on is to produce a machine that takes
codes and produces speach through a speaker. Ideally, it would be a
simple machine, reading pairs of bytes from a serial chip, looking them
up in a table to find a code, which it then writes to a voice synth
chip. Our instructor likes simple, so he approved the project. The
problem is that we're finding it impossible to find a simple voice synth
chip from an era gone past. Nowadays, computers are using
microcontrolers to do voice synthesis.

So, we are exploring the possibility of replacing the synth chip with an
D-to-A converter, and the EPROM with a larger one that contains samples
of phonemes that we make ourselves. The trouble is that, while I have
been studying computers for a very long time, and linguistics long enough
to become reasonably competent, I have never put the two together.

Say I were to sample a few words... first, how would I figure out where
in my sample to make cuts? Second, would the affect of phonemes on their
surroundings make things sound particularly disgusting if I were to put
one into an environment that is slightly different from whence it came?

How about allophones? There are 5 allophones for t in American English:
unaspirated, aspirated, unreleased, alveolar tap, and
retroflex/post-alveolar. Do I need all of them? What would be the
minimum set? How bad would it sound?

How about syllable stress? Should I sample two sets of vowels and
diphthongs, one from positions in words where they are under stress, and
one from where they are not?

Is there a place on the net where I can get a set of phonetic segments
for this kind of voice synthesis? I want at least American (British will
suffice) English, but a few other languages would be nice too. Would I
spend a year trying to figure out what kind of twisted format the samples
are encoded in and how to decode it into what I need to put on EPROM?

I'm trying to do something as quickly and efficiently as possible, but
some elegance and sophistocation would be nice, especially to impress the
class and the instructor.

Thanks very much, and please respond directly to millertgrad.csee.usf.edu.