TMG-L Archives

Archiver > TMG > 2003-05 > 1051964701


From: Michel Vuijlsteke <>
Subject: [TMG] Language strategies
Date: Sat, 3 May 2003 14:27:10 +0200


Hi all,

Is there an accepted TMG strategy for handling multiple languages in your
data?

I've been using TMG for longer than I care to remember, but I've just now
decided to do something with the narrative reports.

Probably about 75% of data I have between 1750 and about 1900 is in French,
with things before or since then are about evenly split between Dutch,
English and French.

This makes for very awkward reports, things like "He was voltigeur au
premier bataillon die 45ieme regiment d'infanterie de ligne" or "She died na
eene kortstondige duizeligheid", which respectively indicate he was a kind
of scout in the Napoleonic army and she died after a short spell of
dizziness.

Is it the accepted practice to translate all these little phrases? Should I
then add the original in parentheses? Or leave that for a footnote?

Am I correct in fearing I'm never going to be able to have good reports in
more than one language? As far as I can see, getting a report to look even
halfway decent involves a *lot* of tweaking sentences, creating short
phrases in memos that don't always mean somehting without the sentence
structure there, creating custom sentence structures for certain
individuals, etc.

As I'm probably most fluent in Dutch I've decided to have a go at fleshing
out the Dutch TMG sentence structures (there's not much there!), but I very
quickly ran into problems I'm not sure how to fix--not necessarily because
it's Dutch, but certainly exacerbated by it.

Take the Address tag. How do you deal with the prepositions for these three
tags? They are for one and the same individual, and i'd like to see them in
one and the same report, so the at/in-switch on the report option / places
tab is no use to me:

* He was born on a farm near Aalter.
* He lived at 4, Crescent Street.
* He lived in Springfield, IL.

Is it the accepted thing to have a default sentence like "[P] lived [L]",
and then manually go in and change it everywhere the generated sentence
doesn't make sense?

In Dutch it's even worse, because to get somewhat readable sentences you
have to add articles, as in:

* Hij woonde op het Dorsplein (he lived on town square, Dutch needs "on
*the* town square")
* Hij studeerde wetenschappen op het Sint-Barbaracollege (he studied science
at CollegeName, Dutch needs "at *the* CollegeName)
* Hij studeerde rechten aan de Rijksuniversiteit (he studied law at
UniversityName, Dutch needs "studied law at *the* UniversityName", also
remark that it's not the same article: DE as opposed to HET)

It doesn't make sense to have all these addresses start with articles, and
except for manually tweaking every single person in a 10.000 person database
I don't see any solution to the problem...

Help!

Michel Vuijlsteke

This thread: