5.0364 Qs: Phonetic French; Mongolian; Sun RDBMS; WP KB (4/132)

Elaine Brennan & Allen Renear (EDITORS@BROWNVM.BITNET)
Fri, 4 Oct 1991 17:49:27 EDT

Humanist Discussion Group, Vol. 5, No. 0364. Friday, 4 Oct 1991.

(1) Date: Wed, 2 Oct 1991 20:43 EDT (58 lines)
From: "Michel (mgrimaud@lucy.wellesley.edu\") GRIMAUD"
Subject: Orthographic to Phonetic French

(2) Date: Thu, 3 Oct 1991 08:26 EDT (23 lines)
From: Jim Wilderotter -- Georgetown Center for Text and
Subject: wordperfect keyboard

(3) Date: Thu, 3 Oct 1991 09:18:18 -0400 (41 lines)
From: rogers@epas.utoronto.ca (Henry Rogers)
Subject: Mongolian query

(4) Date: Wed, 2 Oct 91 11:04 EDT (10 lines)
From: Jean Veronis <VERONIS@vassar.bitnet>
Subject: Q: Relational DBMS for Suns

(1) --------------------------------------------------------------------
Date: Wed, 2 Oct 1991 20:43 EDT
From: "Michel (mgrimaud@lucy.wellesley.edu\") GRIMAUD"
Subject: Orthographic to Phonetic French

I am planning to transcribe about 4,000 lines of French poetry -- the
text of Victor Hugo's masterpiece _Dieu_ (note modest title) -- into
phonetics... and would like the computer to do much of the work for me.

DOES ANYONE KNOW OF A MACHINE READABLE GLOSSARY OF FRENCH WORDS WITH
THEIR PHONETIC EQUIVALENT?

If not I suppose I will have to create it myself. Here is how I imagine
doing it. Please criticize!

Step 1: Scan the text into WordPerfect and proofread

Step 2: run a frequency list so that repeated words do not need to be
transcribed phonetically twice

Up to here things are clear. The next steps are where I would like
suggestions.

Step 3: Find in the linguistic literature... WHERE???... if there are
robust (not fail-safe but useful) sets of correspondences between
spelling and pronunciation in French. (Beyond the obvious.)

Step 4: Find out whether there are robust rules for SYLLABIFICATION
(I remember reading about problems with this... in an
article published 15-20 years ago and read 10 years ago
at least...)
Does there exist a DICTIONARY OF ALL POSSIBLE SYLLABLES?
Would it be of any use?

Step 5: Deal with liaison and enchainement
One solution I envisage is to search for all PAIRS of words and
try to create rules for those pairs that do occur.

Step 6: Deal with the phonetics of end of line to beginning of next line...
Which would be considered performance under ordinary
circumstances but which has to be dealt with here

Step 7: Put all this together

Step 8: Establish links in a database between the text and its
transcription

Any help on any of the steps would be most welcome.

I would, of course, be most curious to know if anyone has already done
this kind of work.

Michel Grimaud
Dept of French
Wellesley College
Wellesley MA 02181

Tel. 617/235-0320 (extension 2404)

(2) --------------------------------------------------------------28----
Date: Thu, 3 Oct 1991 08:26 EDT
From: Jim Wilderotter -- Georgetown Center for Text and
Subject: wordperfect keyboard

Does anyone know where I might find alternate keyboard maps for
WordPerfect? I am especially interested in both Russian (any of the
dialects) and Czechoslovakian.

Any help you can provide would be greatly appreciated.

Sincerely,

James A. Wilderotter II
Project Assistant
Center for Text and Technology
Academic Computer Center
Reiss Science Building, Room 238
Georgetown University
Washington, DC 20057
Tel. (202) 687-6096

BITNET: Wilder@Guvax
Internet: Edu%"Wilder@Guvax.Georgetown.Edu"
(3) --------------------------------------------------------------56----
Date: Thu, 3 Oct 1991 09:18:18 -0400
From: rogers@epas.utoronto.ca (Henry Rogers)
Subject: Mongolian query

Is anyone familiar with writing Mongolian on a Macintosh? I would be
interested in fonts and a suitable operating system. There is a small
group of people here at Toronto working on Mongolian, and we would
appreciate any help.

Mongolian is written with a cursive alphabet in lines running from top
to bottom with the first line starting at the top left and subsequent
lines moving to the right.

The best solution so far seems to be a Hebrew/Arabic word processor with
the letters entered on their side. Naturally Mongolian speakers/writers
find writing sideways odd.The output must be turned 90 degrees to be
read. The right-to-left input puts the lines in the correct order.

A Chinese/Japanese system with vertical input puts the lines in the
wrong order. Further,it is mono-spaced, not suitable for a cursive
writing system.

Does anyone know of a suitable operating system runing from top to
bottom starting at the top left and allowing proportional spacing?

Also, does anyone have a Mongolian font? The only one we have is
bit-mapped and has a rather heavy old-fashioned 19th-century look?

I realise that this problem may seem a touch recherche, but it is
comforting to think that technology is allowing us to think about such
things nowadays.

Thanks,

Henry Rogers
Department of Linguistics
University of Toronto

rogers@epas.utoronto.ca
rogers@epas.toronto.edu

(4) --------------------------------------------------------------15----
Date: Wed, 2 Oct 91 11:04 EDT
From: Jean Veronis <VERONIS@vassar.bitnet>
Subject: Q: Relational DBMS for Suns

Who knows a cheap relational database management system for Sun
workstations? I am not looking for something very sophisticated. Just
basic functionalities.

Thanks,
Jean Veronis