7.0087 Qs: Latin characteristics; quote; s/w recs (3/96)

Elaine Brennan (EDITORS@BROWNVM.BITNET)
Fri, 9 Jul 1993 12:58:10 EDT

Humanist Discussion Group, Vol. 7, No. 0087. Friday, 9 Jul 1993.


(1) Date: Thu, 8 Jul 1993 09:46:52 -0500 (EDT) (55 lines)
From: mccarty@epas.utoronto.ca (W. McCarty)
Subject: Q: characteristics of Latin

(2) Date: Mon, 5 Jul 93 10:56:58 EST (17 lines)
From: "Phyllis Wright" <pwright@spartan.ac.BrockU.CA>
Subject: proverb or quote??

(3) Date: Wed, 7 Jul 1993 09:03:14 -0500 (EDT) (24 lines)
From: mccarty@epas.utoronto.ca (W. McCarty)
Subject: software recommendations

(1) --------------------------------------------------------------------
Date: Thu, 8 Jul 1993 09:46:52 -0500 (EDT)
From: mccarty@epas.utoronto.ca (W. McCarty)
Subject: Q: characteristics of Latin

The following is a query about those characteristics of classical Latin that
one should take into account for automatic processing of the language.
My query has had the benefit of comments from members of LATIN-L@PSUVM.bitnet,
in particular W. Porter (Houston), C. Conrad (Washington U.), R. Coon
(Indiana), H. Stahlke (Ball State), D. Wigtil (US Dept. of Energy), whom I
thank. I'm circulating the updated query here in hopes that other Latinists
and linguists will speak up. I'd be grateful for any bibliographical items
closely related to the subject.

-----

My question is this: what aspects of the language, including commonly
exercised 'poetic license', might affect computer-based analysis of poetry
written in classical Latin, especially Ovid? In particular, what phenomena
would cause a purely automatic analysis of an unencoded text to err and so
serve as an argument for encoding? Since I only know a few languages and am
not a linguist properly speaking, I'm somewhat hesitant to rank Latin as being
esp. challenging in certain respects. Nevertheless I must. Hence I'd very
much appreciate comments from members of this group on these broadly
comparative assertions:

1. Latin is highly inflected, often irregularly.
2. Relations between words are signified almost exclusively by
grammatical concord or agreement rather than by word-order.
3. Word-order is very fluid, especially in poetry, where
syntactically related words are often widely separated.
4. Latin has a fully developed system of gender (m,f,n), which
is extensively used to mark related words.
5. As a rule, the expressive power and subtlety of Latin are based on
a relatively small active vocabularly of words with comparatively
broad and diverse semantic fields. (Should roots with multiple
meanings be treated as polysemous or synonymous?)
6. It has a large number of homographs.
7. The notion of 'sentence' (in our sense) is not well defined. (To what
extent are other structural units comparatively ambiguous?)
8. Plural nouns, pronouns, nominal adjectives, and related verbs are often
used for singular subjects (i.e. where in English we would expect
the singular).

Are there other characteristics I should be taking into
consideration? Why would you find fully automatic techniques of
analysis (e.g., text retrieval by keywords, proximity tests) esp.
difficult with Latin?

Thanks very much. Replies to the entire group are probably best. And please
circulate this note wherever it might provoke some reply.


Willard McCarty
mccarty@epas.utoronto.ca

(2) --------------------------------------------------------------34----
Date: Mon, 5 Jul 93 10:56:58 EST
From: "Phyllis Wright" <pwright@spartan.ac.BrockU.CA>
Subject: proverb or quote??

I have been trying to discover where the following appears:

"We understand ourselves when we understand our parents."

I have checked all the quotation/proverb books in my library.
Can someone help?

Phyllis Wright
Phyllis M. Wright (416)688-5550, ext. 3961
Brock University Library pwright@spartan.ac.brocku.ca
St. Catharines, Ontario
Canada L2S 3A1

(3) --------------------------------------------------------------42----
Date: Wed, 7 Jul 1993 09:03:14 -0500 (EDT)
From: mccarty@epas.utoronto.ca (W. McCarty)
Subject: software recommendations

I am in need of wisdom from experience with the following commercial
software (all for MS-Windows):

Paradox
Access
Excel
Quattro-Pro

I need one relational dbms and one spreadsheet package to use in a
humanities computing course at the graduate (but elementary) level.
I'd be very interested in comments as to which are excellent in
themselves and which dbms works better with which spreadsheet. If
there are questions I have forgotten to ask, then answers to these
would also be welcome.

Thanks very much.

Willard McCarty
mccarty@epas.utoronto.ca