Forum .LRN Q&A: Support of professional translators
Therefore I also could not resist to translate many
messages that should rather have been left to the
professional translators from places like a
university's Institute of Translating and Interpreting.
Are there such professional volunteers already in
this forum?
Since I do not yet know enough about the environment of
this project, another question: does anybody know of
if there are connections to the I18N project of the
KDE translating group, and if the terminology database
there under the address
https://userbase.kde.org/Glossary
is really the current one, or if it is given up in
favor for
https://web.archive.org/web/20170321112445/http://opensource.bureau-cornavin.com/glossary/
or if another one can be used?
Working with the great Wysiwyg translator interface, I
noticed following aspects:
1. Interdependencies
The problem of homogeneity (which was also mentioned
by the German KDE translators,
http://i18n.kde.org/teams/de/#einheitlichkeit ),
is really very important. It must not occur that
"My Space" is "Meine Startseite" on one page and
"Eigene Startseite" on the other. But how can these
interdependencies be established? I think, at least
- all (e. g.) German translations of a given English
word should be able to be displayed on one query,
- all English originals leading to a given German word
should also be query-able, and
- ideally, also semantically similar wordings should be
able to be grouped together, and the same should apply
to similar roles of a wordings within a menu, e. g.,
first mention (more verbose), vs. shortcut or option
list (less verbose).
Of course, with XML all the various interdependencies
could be perfectly modeled and automatically processed.
But this would probably be a too ambitious project
which would never finish. Therefore, a more appropriate
XML structure might be one that facilitates the manual
processing of a semantic field in parallel to the
fine new interface already finished.
An approach that seems promising to me is XTM for
the semantic web (XSW, see Beisswenger/Lenz/Storrer
"terminologiebedingten Wissensvoraussetzungen" in this
publication list
https://web.archive.org/web/20040119223002/http://www.hrz.uni-dortmund.de:80/~hytex/hytex/worx.html.)
2. Retrieval
The most important messages are easy to be discovered
and processed with the pretty Wysiwyg interface (and this
is good because it is them that the user must most
frequently work with). But the less frequent messages
and those occurring only under very special circumstances
are much harder to be reached in context and on the
normal click path. Therefore, the temptation is great
to translate them via the message list, without really
understanding what exactly happens in this context
(and that's exactly what an external translator does
who does not understand what s/he is doing, anyway).
The display of ""Message key usage" (like /web/translate/
packages/dotlrn/dotlrn.info) does not sufficiently solve
this problem. Instead, it would be desirable to somehow
map these cryptic module names (understood only by the
programmer himself) onto the menus that could be
retrieved in a sitemap and could be translated into
a click path.
Above all, the messages should be search-able, by English
and (e. g.) German words/strings, ideally by additional
attributes of a terminology database.
3. Linguistic nuances
Often the greatest problem is to leave the focus of a
menu area to stay as easy to catch (cognitive economy
and ergonomics) as it was in En_US. Example:
- "Delete folder" (focus on the predicate) becomes either
- "Lösche Ordner" (very terse), or otherwise, one gets
plenty of problems with clarity, politeness, length,
directness/distancelessness or unpersonal style:
- "Ich" (das System) "lösche den Ordner" ("I" (the
system) "am deleting the folder")
- "Lösch" (Du/du) "den Ordner" (imperative, second
person singular, i. e., the problematic familiar
"Du" (in old orthography) or "du" (old orthography for
schoolbook-like instructions and according to the
new orthography now a general rule but one of the most
controverse instances of the new orthography)
- "Löschen Sie den Ordner" (formal variant of imperative)
- "Ordner wird gelöscht": unpersonal passive form
("folder is being deleted"), which has again the wrong
focus, but the corresponding unpersonal active
form is too unusual "Man löscht den Ordner" ("one is
deleting the folder" = "you are deleting the folder" or
even antiquated ("Man lösche den Ordner" ~ "thou
shalt delete the folder").
I think, the professional translators are needed here. And
therefore I unrestrainedly experimented with various
wordings because I think its just a preliminary first
version, anyway.
About linguistic nuances:
You are right, there are some translation problem from different language environments, e.g. from Englisch to German, because of different phrase structures. But I do not see it worth making long reflections about every item. There is enough experience in the software world about translating phrases like "delete folder". Here you normally have just an inversion of the "focus", so: "Ordner löschen" is the most commonly chosen translation in the software world. I would just stick to already existing conventions as far as it is possible.
The real problem that I see here is another one:
CONSISTENCY
Take, for example the word ITEM. In Italian I have seen it translated as ELEMENTO. Though in the .LRN evironment ITEM actually refers to "content item". So it could as well been translated as CONTENUTO.
The same is for German ITEM = ELEMENT or INHALT
Im I right???
How can we assure consistency in the translation work?
I talked with Peter and Lars from Collaboraid about it and I get the feeling that this could be solved by retrieving similar words and phrases from the source and target languages and presenting them in the translator's UI (as Matthias mentions above).
Some other things that came up in the discussions was a way to improve communicating the context of used terms as well as access to translation glossaries within the interface.
Examples of glossaries we ran across that might be of value to others:
Microsoft's software glossaries (translations that tend to become convention):
ftp://ftp.microsoft.com/developr/MSDN/NewUp/Glossary/
Frank Dietz's collection:
https://web.archive.org/web/20040611140150/http://www.jump.net:80/~fdietz/
Once I get a chance to touch base with Lars I will post a ToDo list. Regretfully, there is still some basic i18n infrastructure that needs to be finished before we can make this fun stuff happen.
