Poll: Do you pre-translate documents using your CAT tool's termbase? Thread poster: ProZ.com Staff
|
|---|
This forum topic is for the discussion of the poll question "Do you pre-translate documents using your CAT tool's termbase?".
This poll was originally submitted by Miguel Llorens
View the poll here
A forum topic will appear each time a new poll is run. For m... See more This forum topic is for the discussion of the poll question "Do you pre-translate documents using your CAT tool's termbase?".
This poll was originally submitted by Miguel Llorens
View the poll here
A forum topic will appear each time a new poll is run. For more information, see: http://proz.com/topic/33629 ▲ Collapse
| | | |
Reed James
Chile
Local time: 16:00
Member (2005)
Spanish to English
| I never do it. | May 28, 2006 |
I tried pre-translating with both the Translation Memory and Term Base when I was testing CAT software like Word Fast and Deja Vu. I find that the pretranslated bits get in the way of the final translation. The translator runs the risk of leaving the words in the wrong place or including them twice.
In short, it is more work than it is a help.
Reed D. James
| | | |
Andrey Belousov (X)
United States
Local time: 16:00
Russian to English
+ ...
| Do not have brains? - USE IT! | May 28, 2006 |
Hi Reed! Never do I! Loss of effort - I trust my head way more than any program/ I'm with you on that!
| | | |
| Almost always pretranslate with DVX | May 29, 2006 |
Because of the clearly structured DVX interface I find this helpful. And the way DVX assembles segments from its various databases means that the results (in terms of words and phrases) are meaningful and helpful. And I even set DVX to put words from the source text into the assembled target text if there is nothing in the databases (this helps me not to inadvertently forget bits of sentences).
However, I work in languages with very different sentence structures (GermanEnglish), so I often... See more Because of the clearly structured DVX interface I find this helpful. And the way DVX assembles segments from its various databases means that the results (in terms of words and phrases) are meaningful and helpful. And I even set DVX to put words from the source text into the assembled target text if there is nothing in the databases (this helps me not to inadvertently forget bits of sentences).
However, I work in languages with very different sentence structures (GermanEnglish), so I often need to reshuffle the word order, rephrase things, choose different terminology (either different options from my databases or terms that are not yet in the database). And the flow of the sentence often needs rephrasing anyway (e.g. German noun phrases need to be transposed into English verb phrases), so the pretranslated/assembled material from my databases is often just a starting point.
For me, I find that the pretranslate feature (along with other CAT features in DVX) helps me to produce better work. In the "bad old days" of typing to Word from paper or overwriting texts in Word, I sometimes missed sentences or paragraphs, had more typos than I have now - and now I find it easier to concentrate on the content and wording, and to proofread and check for plausibility as I go along.
The style of working is very individual and personal. My work process would not work for everyone, and I know that there are DVX users who go about things differently.
Miguel, thanks for the interesting poll topic. I look forward to seeing how the results unfold. ▲ Collapse
| | |
|
|
|
I tried it with Trados and it doesn't help at all - I think that lone words in a untranslated sentence make a mess more than anything else.
The Assemble function of DejaVuX sometimes takes isolated words from the termbase but the layout makes it easier to avoid errors - Ctrl+Space simply erases the pretranslated sentence if I prefer to start from scratch.
OTOH, I use the termbase quite intensively during translation.
| | | |
| So do you add up in your head rather than create a formula in Excel? | May 29, 2006 |
Andrey Belousov wrote:
Hi Reed! Never do I! Loss of effort - I trust my head way more than any program/ I'm with you on that!
I wonder whether in Excel you'd insist on adding a column up manually rather than using a formula? Making use of the TermBase (which does not involve a machine translating for you) does not imply not having brains.
| | | |
Ian M-H (X)
United States
Local time: 16:00
German to English
+ ...
| Can be good tools for terms ;-) | May 29, 2006 |
Maybe the question's phrasing ("pre-translate documents") has led to a couple of very firm negative responses, and of course not every CAT tool has something it calls a "termbase".
Andrey wrote:
I trust my head way more than any program
Well, yes, but no one suggested switching your brain off and not monitoring/checking any substitutions made by the terminology database.
and Marie-Céline wrote:
ne words in a untranslated sentence make a mess more than anything else.[/quote]
It can do, but I think this depends on the text and whether or not we're really talking about terms. If I'm working on a piece of political journalism, neither a translation memory nor a terminology database is likely to help me very much with 99% of the text. The terminology database might save me looking up the name of a ministry or a European institution, but that's about it.
But if I'm translating a catalogue, getting client-specific terminology (1) right and (2) out of the way can free up my brain for the more challenging bits. If I'm constantly trying to remember whether Company A calls its products baths, washbasins and vanity units (as opposed to Company B's bathtubs, wash stands and base units), I'm less likely to do justice to their distinctive lines, sweeping curves, timeless elegance or whatever. Multiply that effect by 20, to take account of the other 60 or so pieces of kit for which the client has established translations, and there are real benefits to be had in terms of both time and quality.
[Edited at 2006-05-29 07:50]
| | | |
bobedwin 
Local time: 22:00
Italian to English
+ ...
I was happy to see that I am not the only one that does not have a CAT tool and that the numebr of colleagues like me is (though a minority) more than I expected.
I started at the times of typing machines and I have not yet the courage to out my nose into CATs, moreso I have not yet had a dire need of using them.
I generally translate straight off and then take my time to re-read and polish the translation.
| | |
|
|
|
| Assemble from portions with DVX | May 29, 2006 |
I agree with Victor.
IMHO, DVX' best feature is Assemble from Portions - especially for technical texts. And, nobody forces me to accept the assembled segment, if it does not fit I just do Copy source to target ...
Interesting topic, maybe for one of the next conferences we could organize a workshop for long-time CAT tool users to "compare experiences/techniques".
Victor Dewsbery wrote:
Because of the clearly structured DVX interface I find this helpful. And the way DVX assembles segments from its various databases means that the results (in terms of words and phrases) are meaningful and helpful. And I even set DVX to put words from the source text into the assembled target text if there is nothing in the databases (this helps me not to inadvertently forget bits of sentences).
However, I work in languages with very different sentence structures (GermanEnglish), so I often need to reshuffle the word order, rephrase things, choose different terminology (either different options from my databases or terms that are not yet in the database). And the flow of the sentence often needs rephrasing anyway (e.g. German noun phrases need to be transposed into English verb phrases), so the pretranslated/assembled material from my databases is often just a starting point.
For me, I find that the pretranslate feature (along with other CAT features in DVX) helps me to produce better work. In the "bad old days" of typing to Word from paper or overwriting texts in Word, I sometimes missed sentences or paragraphs, had more typos than I have now - and now I find it easier to concentrate on the content and wording, and to proofread and check for plausibility as I go along.
The style of working is very individual and personal. My work process would not work for everyone, and I know that there are DVX users who go about things differently.
Miguel, thanks for the interesting poll topic. I look forward to seeing how the results unfold.
| | | |
Vito Smolej
Germany
Local time: 22:00
Member (2004)
English to Slovenian
+ ...
SITE LOCALIZER | Being among those negligible 1.x percent who always do | May 29, 2006 |
Why do I do it?
It has something to do with the subject or subjects one is acquiring: let's say you earn your living with localizing prescriptions. After a while you SHOULD hate to translate again and again the same sentences. The same applies to user manuals, safety data sheets, marketing material...
Second case for pretranslation: price lists need to be updated from 2005 to 2006. The customer needs installation manual for the other machine. The maintainance manual i... See more Why do I do it?
It has something to do with the subject or subjects one is acquiring: let's say you earn your living with localizing prescriptions. After a while you SHOULD hate to translate again and again the same sentences. The same applies to user manuals, safety data sheets, marketing material...
Second case for pretranslation: price lists need to be updated from 2005 to 2006. The customer needs installation manual for the other machine. The maintainance manual is now outstanding. etc etc. etc.
So actually I must wonder about the translations those 47%, who never pretranslate, are doing.
One more point: Pretranslation ( if the agency does not squeeze you) is an easy way to spend less time per buck paid (i.e. earn more per hour). It just takes a large enough translation memory.
Im using TRADOS, but the points pertain to any half-decent TM-based CAT tool.
smo ▲ Collapse
| | | |
| It depends enormously on the type of text... | May 29, 2006 |
... and whether there is anything useful in the TM at all!
I have some regular menu clients. It's fun to watch a 200-word menu translate itself, but it doesn't save a lot of time, and I sometimes scrap the Trados version and start again 'the old-fashioned way'. On other occasions the concordance is still a great help.
I've just done a really big job - for me - an 8000-word contract, with 11 files of attachments and technical conditions.
Trados was more trouble tha... See more ... and whether there is anything useful in the TM at all!
I have some regular menu clients. It's fun to watch a 200-word menu translate itself, but it doesn't save a lot of time, and I sometimes scrap the Trados version and start again 'the old-fashioned way'. On other occasions the concordance is still a great help.
I've just done a really big job - for me - an 8000-word contract, with 11 files of attachments and technical conditions.
Trados was more trouble than it was worth - it found fuzzy matches everywhere that I had to spend ages proofreading and re-translating. In the end I just inserted the source text, and for most files gave up on Trados altogether.
I have learnt to like Trados for the jobs where it helps, but there are still plenty where it is simply a waste of time. There are no matches, and never will be, however long you carry on extending the TM.
A potato peeler is great for peeling potatoes, but it's useless for whisking eggs. There are translators who specialise in the jobs where CATs are brilliant, and there are jobs where a CAT is useless.
The great thing is that there is always someone out there who loves the jobs I hate, and others who barely dare start on the jobs I think are fun!
Vive la différence and happy translating, folks! ▲ Collapse
| | | |
| DV Autosearch | May 29, 2006 |
I don't pretranslate, but I do always work with DV's AutoSearch windows open. That way all terms in the source that have entries in my termbase are visible (as well as any fuzzy or exact sentence matches), and if I want to use them in the translation, I can insert them with a simple two-key command.
I find it more convenient and functional than using a pretranslation function; the target isn't cluttered up with matches, yet they are all visible and available, and I can choose any t... See more I don't pretranslate, but I do always work with DV's AutoSearch windows open. That way all terms in the source that have entries in my termbase are visible (as well as any fuzzy or exact sentence matches), and if I want to use them in the translation, I can insert them with a simple two-key command.
I find it more convenient and functional than using a pretranslation function; the target isn't cluttered up with matches, yet they are all visible and available, and I can choose any that I want to use and insert them quickly and simply. ▲ Collapse
| | |
|
|
|
Ilde Grimaldi
Local time: 22:00
Member (2003)
English to Italian
+ ...
| assemble with DVX | May 29, 2006 |
I agree with all message by Victor Dewsbery.
I can add that I would do it even if I had to translate general texts, because they anyway would include usual terms, for example in English "and", "for", "with", "June", "yellow" etc.
Translating technical texts, for example online help in IT documentation, I would hate to always overwrite "installation", "files", "dialog box", "double-click the", "for more information, see" etc.
Please note that I'm not talking about... See more I agree with all message by Victor Dewsbery.
I can add that I would do it even if I had to translate general texts, because they anyway would include usual terms, for example in English "and", "for", "with", "June", "yellow" etc.
Translating technical texts, for example online help in IT documentation, I would hate to always overwrite "installation", "files", "dialog box", "double-click the", "for more information, see" etc.
Please note that I'm not talking about pretranslating at the beginning of a document, but about assembling each and every new segment to translate (when you close a segment, the next one gets assembled). In DVX, it is very simple to add terms or syntagms on the fly, so when you assemble the new segment you use all terminoly added until that moment, for example in technical documents that would be "combustion chamber shut-off valves", "will be partially offset by" (in financial statemens) etc.
Another good thing about assembing in DVX (:-)) is the fact that not only terms from 2 glossaries plus 1 lexicon are used to get an assembled sentence, but also whole segments from the memory, which could be elements of this current bigger sentece.
I too have set the option to insert the unfound words, this is way for me to avoiding for sure to overlook some term of the source text.
I've been using this method for years, delivering documentation mostly in Trados format, and I could only work this way
[Edited at 2006-05-29 17:25] ▲ Collapse
| | | |
Login to reply/comment 
