Re. More notes on translation

Subject: Re. More notes on translation
From: Geoff Hart <geoff-h -at- MTL -dot- FERIC -dot- CA>
Date: Thu, 21 Dec 1995 10:49:03 LCL

Diane Ritchey suggested that she had been advised by
various folks at "translation houses" that "translations
take roughly as long as the original writing". Not to put
too fine a point on it, Diane, but they're either
intentionally deceiving you to guarantee that they've got
more time than they need to meet your deadline, or they're
not particularly skilled at their jobs.

I do French to English translations of all our technical
reports (forestry engineering, reforestation, etc.) as part
of my job. I'm pretty good at it, but I'm not Mr.
Translation. I can translate a 32-page report complete with
graphics, tables and figures in about two days, including
interruptions and bugging the author for clarification;
that's probably about 10 pages of 8.5 X 11 single-spaced
text per day after you delete the visual parts. If the text
is much more complicated, maybe three days (say 7 pages per
day). Revision and approval takes about another 2 days all
told, but broken into several chunks because the manuscript
has to circulate. Don't ask how long it took the author to
write it (including research and data analysis), me to edit
it (before translation), and the peer review process to
spit it back out again so we could publish it... just take
my word for it that it's significantly longer than four
days. <grin>

Our professional French translator (thanks, Therese!), who
does this work full-time, is at least as fast (often
faster), and she doesn't get to bang on an author's door
with her questions. Despite relying on e-mail and phone
calls, her work also requires less revision than mine. (On
the other hand, the problem isn't with my technical
accuracy; because I'm the editor, and cover their pages
with red ink, my authors take a certain pleasure in editing
_me_ and playing with _my_ wording. Go figure! <grin>)

It is possible that your translation houses are including a
full review and usability test of the documentation
(described below), including whatever review cycles you
would practice for the English version, in which case their
estimates are closer to the truth... but it still won't
take the same length of time to translate as to write. Not
even close. Think about it: how much work do you do just to
collect the information you need to begin writing? Do they
do this work all over again in the second language? I doubt
it. How much revision do you do until you get the thoughts
and the words that express them just right? You've already
done all that work for the translater.

Your second question concerned testing the quality of a
translation. There are always two main tests you'll have to
do: test with your users, and test with your technical
experts. The users must read the translation to verify that
the language is appropriate for their needs; for example,
when I worked for the federal government, our French
translaters were accurate, but they often used fancy words
that were trendy (say) in Paris or some literary magazine,
but our readers had never heard of these words. (In some
cases, the words weren't even in our basic dictionary!) I
always simplified these words; I have this unreasonable
philosophy that if a sample of your audience doesn't know a
word, they should at least be able to find it in the sort
of dictionary a non-editor would own. <grin>

The techies must read the translation to be sure that the
correct words were used; there are lots of near misses in
technical translation, and only an expert who also speaks
the language fluently will be able to tell you if the
correct word was used. If you don't have an expert who's
skilled in the second language, you'll have to find one.
The other task, particularly if you don't understand the
language yourself, is to make sure that every sentence in
one version was transferred into the second version, and
that every number, cross-reference, name, etc. was retyped
correctly. Try your local STC chapter, or one of the
European chapters, to find a good bilingual editor. Some
technical editors can verify that everything was retyped
correctly _and_ that the translater used the correct jargon
_and_ that it reads well for the reader, but you'll still
want to test things on your audience.

Alexander von Obert, if you're listening, can you provide
suggestions for European contacts in editing and localizing
second-language versions of an English text? Can anyone
else provide Asian (etc.) contacts? I'm thinking of
addresses for STC in other countries, or equivalent
organisations (e.g., translation associations such as ATA).

That's the short version; I'll elaborate further if you're
interested. I'm also sure that this message will attract
some contrary opinions, which I'm looking forward to

--Geoff Hart @8^{)}
geoff-h -at- mtl -dot- feric -dot- ca

Disclaimer: If I didn't commit it in print in one of our
reports, it don't represent FERIC's opinion.

Previous by Author: Re. (Groan...) Gender neutral pronouns
Next by Author: The graphistes strike back!
Previous by Thread: SGML & Databases at the January SGML Forum of NY
Next by Thread: Re. More notes on translation

What this post helpful? Share it with friends and colleagues:

Sponsored Ads