RE: Translations... realism preserves the reader experience

Subject: RE: Translations... realism preserves the reader experience
From: Susan Harkus <Susan -dot- Harkus -at- xt3 -dot- com -dot- au>
To: 'Max Wyss' <prodok -at- prodok -dot- ch>
Date: Tue, 3 Oct 2000 12:19:21 +1100

Max,
Point me to your research.

Mine is at http://www.multilingualwebmaster.com/library/writing-TR.html
Open the PowerPoint slide and read the notes.

I tried very hard to prove that there was nothing to be leveraged from translation software and I was only using the cheap version of Systran (= Systran Classic) rather than Systran Professional... BUT the results to the contrary were repeated over and over again.

I corresponded with a number of experienced people in this field, and followed their research pointers to identify an evaluation strategy that was robust and believable. There is a ton of research on evaluating translations.

The French native speaker with whom I worked took very little time to brush up the draft software translations of the test documents. I am sure that a professional translator would have done so in a shorter time.

The secret is good writing and a flexible dictionary in the translation software. The dictionary trains the software to interpret text strings with specific meanings and syntax. Have you tried customising a dictionary with the translation software that you have evaluated or have you only tried web translations that have no user-defined dictionary?

I know that I can leverage translation software by developing good quality source, especially when the readers tell me they can't see any difference in writing style. I have, of course, modified my writing style slightly to compensate for human and software brains that need to parse text. But in the end, that simply provides extra support for the human readers.

You may not think you can leverage anything from translation software. That's fine. But the option that you are advising people to reject can save a lot of money and get product to the target market in a shorter timeframe.

I don't want to convince you, Max. I am just anxious that others keep their options open so if we have more to discuss on this topic, let's do it directly off-list.
Susan Harkus
Sydney, OZ





Previous by Author: Translations... dogma or realism?
Next by Author: RE: Questions for usability surveys
Previous by Thread: Document Creation Tools
Next by Thread: Name that TLA


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


Sponsored Ads