| Pages in topic: [1 2] > |
Do CAT tools worsen translator's memory and understanding? Thread poster: sjmdcl (X)
|
|---|
sjmdcl (X)
Local time: 05:08
English to Spanish
+ ...
Hello,
I'm currently translating an assignment using a CAT tool. The thing is that I've noticed that my understanding of the whole document decreases and I've found myself translating sentences which don't make sense at all. On the other hand, when I translate without CAT tools I usually learn a lot and have more understanding of the documents so I can improve its style. What are your thoughts?
Greetings.
[Edited at 2012-01-25 15:42 GMT]
[Edited at 2012-... See more Hello,
I'm currently translating an assignment using a CAT tool. The thing is that I've noticed that my understanding of the whole document decreases and I've found myself translating sentences which don't make sense at all. On the other hand, when I translate without CAT tools I usually learn a lot and have more understanding of the documents so I can improve its style. What are your thoughts?
Greetings.
[Edited at 2012-01-25 15:42 GMT]
[Edited at 2012-01-25 15:43 GMT]
[Edited at 2012-01-25 15:43 GMT]
[Edited at 2012-01-25 15:44 GMT] ▲ Collapse
| | | |
S P Willcock (X) 
Local time: 12:08
German to English
+ ...
| depends, I suppose, on how familiar you are with their use. | Jan 25, 2012 |
I have recently been asked to acquire a CAT tool by my main commericial client, and complied because they accounted for very nearly half my income last year. Literary and academic clients are more fun, but less money, so I recognised the need to buy the tool. I bought it yesterday and used it for the first time today.
I can imagine, with several weeks' training and months of practice, reaching a stage where it "fades into the background" and I can work fluently with the tool, but r... See more I have recently been asked to acquire a CAT tool by my main commericial client, and complied because they accounted for very nearly half my income last year. Literary and academic clients are more fun, but less money, so I recognised the need to buy the tool. I bought it yesterday and used it for the first time today.
I can imagine, with several weeks' training and months of practice, reaching a stage where it "fades into the background" and I can work fluently with the tool, but right now I can definitely state that my understanding of source text and my fluent output of target both suffer while I wrestle with the blasted software.
Tangentially, I watched some training videos from the manufacturer earlier this evening. One of the videos is a walkthrough on how to use the tool to translate a document EN>GER, and whenever the presenter reads out a German word to make a point about how the software works, it is evident that she neither speaks nor understands German.
Why did the manufacturer choose her to front the video? Is the woman a translator at all? What languages does she work with? Could she not have walked us through a document in her working languages? I know that this is a small point but it seems to me symptomatic of some of the problems of CAT tools - that they are used as a substitute, rather than a supplement, for proficiency. ▲ Collapse
| | | |
| My thoughts too | Jan 25, 2012 |
And it gets worse when the punctuation rules of the target differ from those of the source, so you end up getting say half sentences all over the place (or sometimes segments in the wrong order, etc.). And no, you can’t fix it by changing the segmentation rules, or you end up getting the opposite problem, entire paragraphs as segments, which is even worse.
Things do seem to get better with use, but these problems do persist. Sometimes you can’t even merge the offending segments ... See more And it gets worse when the punctuation rules of the target differ from those of the source, so you end up getting say half sentences all over the place (or sometimes segments in the wrong order, etc.). And no, you can’t fix it by changing the segmentation rules, or you end up getting the opposite problem, entire paragraphs as segments, which is even worse.
Things do seem to get better with use, but these problems do persist. Sometimes you can’t even merge the offending segments because the CAT thinks that they are different paragraphs… ▲ Collapse
| | | |
LEXpert
United States
Local time: 04:08
Member (2008)
Croatian to English
+ ...
| Probably not a translator at all | Jan 25, 2012 |
S P Willcocks wrote:
Tangentially, I watched some training videos from the manufacturer earlier this evening. One of the videos is a walkthrough on how to use the tool to translate a document EN>GER, and whenever the presenter reads out a German word to make a point about how the software works, it is evident that she neither speaks nor understands German.
Why did the manufacturer choose her to front the video? Is the woman a translator at all? What languages does she work with? Could she not have walked us through a document in her working languages?
She's probably not a translator at all, but most likely a professional trainer for the company, or one of their staffers who routinely does training for them. If it was a general training video aimed at all users, they probably just picked EN>DE as a common example (after all, they had to pick something), and there are hundreds or thousands of potential combinations among the audience of all translators who might need the software, so there would be no way to cover them all.
I been through all sorts of trainings when I had a corporate job - and in most cases, even if the training was customized, the trainer has only a basic idea of your specific application. If it's a generalized training, they probably have no idea at all, unless you asked specific questions yourself.
Unless the video was specifically directed at DE-EN translators, there's nothing unusual in the scenario you describe. The mechanics of the tool are the same regardless of the language combination (perhaps with the exceptions of pictographic or differently oriented languages), and any explanations she would have provided about specific DE-EN issues would have been irrelevant for viewers who don't know those languages.
S P Willcocks wrote:
I know that this is a small point but it seems to me symptomatic of some of the problems of CAT tools - that they are used as a substitute, rather than a supplement, for proficiency.
To me, that proves that opposite - CAT tools can potentially make anyone a faster translator (if not necessarily a richer one!), but not a better translator.
| | |
|
|
|
B D Finch
France
Local time: 11:08
French to English
+ ...
| It depends what CAT tool you use and how | Jan 25, 2012 |
I certainly would think it a danger sign if I felt that I didn't understand the document I was translating. With large and complex documents, I generally print out the source file and read it through on paper before starting to translate. I mark up any points that I need to clarify, vocabulary and translation problems that I notice at that stage. Particularly if I think that there are likely to be segmentation problems and if the format is not over complex (i.e. not full of textboxes, graphics w... See more I certainly would think it a danger sign if I felt that I didn't understand the document I was translating. With large and complex documents, I generally print out the source file and read it through on paper before starting to translate. I mark up any points that I need to clarify, vocabulary and translation problems that I notice at that stage. Particularly if I think that there are likely to be segmentation problems and if the format is not over complex (i.e. not full of textboxes, graphics with labels etc.), I prefer to translate using Wordfast Classic, which allows me to amend any awkward segmenting as I go along.
I also often like to use two screens with the source document displayed on the spare screen, so that I can quickly check against it if needed. The advantage over paper is that allows a quick search to find any text that is not in its normal position when using Wordfast Pro.
I wouldn't be without a CAT tool, but a tool that is a hindrance rather than a help is either being used incorrectly, or should be changed for a different one. ▲ Collapse
| | | |
esperantisto
Local time: 12:08
Member (2006)
English to Russian
+ ...
SITE LOCALIZER | Habits rather than skills | Jan 26, 2012 |
I don’t think, CATs worsen your memory or skills. But they may worsen your habits. Indeed, you get to think in therms of segments rather than documents and stick to kinda ‘translate-and-forget’ approach. When you translate a 100-th operating manual for, say, a pipe bender, this is no harm, but when you plunge into a new area, well, read the whole document first.
| | | |
Giles Watson
Italy
Local time: 11:08
Italian to English
In memoriam
esperantisto wrote:
but when you plunge into a new area, well, read the whole document first.
And read it - preferably out loud - afterwards as well to make sure you have nailed the appropriate rhythms in the target language.
| | | |
neilmac
Spain
Local time: 11:08
Spanish to English
+ ...
| Swings and roundabouts | Jan 26, 2012 |
Rudolf Vedo CT wrote:
CAT tools can potentially make anyone a faster translator (if not necessarily a richer one!), but not a better translator.
Exactly! CAT tools, particularly one whose name I dare not mention, certainly seems to make agencies and middle-fiddlers richer and more demanding though.
Whether it affects your memory is moot and depends on many factors. I suppose that since it means you no longer have to rely on human memory so much, it may seem to "worsen" recall, but having TMs and glossaries means that you don't HAVE to depend on your own memory to the same degree as before, so you don't need to worry so much about forgetting things. For example, I keep having to check the proz glossary for odd terms that I've entered or asked myself in the past; I know that these items are in there, so I don't need to lug them around like excess baggage in what passes for the "active" part of my poor old brain...
Similar concerns have been expressed about googling... but hey - the world in general seems to be going to hell in a handcart anyway - so who's counting?
| | |
|
|
|
neilmac
Spain
Local time: 11:08
Spanish to English
+ ...
| Not an issue | Jan 26, 2012 |
B D Finch wrote:
I certainly would think it a danger sign if I felt that I didn't understand the document I was translating.
I often translate documents without reading them through first, or attempting to understand them, although I do like to have a quick look at it to check the format is OK. And I also translate or revise some stuff (fuzzy logic and algorithms and things like that, or deep brain surgery) I have no idea about. None whatsoever. It is all gibberish to me, but apparently it's well translated gibberish when it's published.
| | | |
Theo Bernards (X)
France
Local time: 11:08
English to Dutch
+ ...
| Wow! Your kidding, I hope? | Jan 26, 2012 |
neilmac wrote:
... And I also translate or revise some stuff (fuzzy logic and algorithms and things like that, or deep brain surgery) I have no idea about. None whatsoever. It is all gibberish to me, but apparently it's well translated gibberish when it's published.
What is the use of having specialisations if one can accept assignments that are gibberish to oneself?
To go back to the original question: no, I don't think that CAT-tools worsen translators memory and understanding, just as power tools don't make carpenters forget how to saw with a hand saw. The most important tool you have is the one between your ears, all other tools are just, well, means to work faster/better/smarter (at least, they should do).
| | | |
| it is just a tool | Jan 26, 2012 |
I use OmegaT because it is free, works on a Mac and I wanted to try out working with a CAT.
Some features are very useful.
The segmentation ensures that you don't skip something.
You can search words in both target and source.
You always know where you are in the text.
*Possible to make a glossary as you go
*Can integrate dictionaries and glossaries.
*These are the star features for me as they save a huge amount of time.
My fi... See more I use OmegaT because it is free, works on a Mac and I wanted to try out working with a CAT.
Some features are very useful.
The segmentation ensures that you don't skip something.
You can search words in both target and source.
You always know where you are in the text.
*Possible to make a glossary as you go
*Can integrate dictionaries and glossaries.
*These are the star features for me as they save a huge amount of time.
My fields have little use for translation memories so I can't comment on that feature. However I find the Google Translate only useful for suggesting some words or turns of phrase (my memory is not what it was). The suggested sentence nearly always has to be rewritten and corrected. As I can't get the new Google Translate API key activated, I've been working without that feature for a while now and find that I work faster without it, even with having to check more words.
It is a useful tool for me. I can imagine that people who don't have strong writing skills in the target language, and who rely heavily on what Google Translate suggests, may be at a disadvantage as they have to think more about each suggestion. ▲ Collapse
| | | |
Kirsten Bodart
United Kingdom
Local time: 11:08
Dutch to English
+ ...
| Exactly my thoughts! | Jan 26, 2012 |
I use a CAT for one client because he wants a TM (God knows why, there is no Dutch person who ever writes the same sentence twice...) and if the file seems to contain a lot of repetition (think questionnaire or technical), but...
The problem is that a CAT severely inhibits your 'text' view in favour of a 'sentence' view. You start to work per sentence. For a technical text that is not exactly a problem, but for something else that approach makes for a rambling piece of language, at ... See more I use a CAT for one client because he wants a TM (God knows why, there is no Dutch person who ever writes the same sentence twice...) and if the file seems to contain a lot of repetition (think questionnaire or technical), but...
The problem is that a CAT severely inhibits your 'text' view in favour of a 'sentence' view. You start to work per sentence. For a technical text that is not exactly a problem, but for something else that approach makes for a rambling piece of language, at best.
From German, that essentially means, working per part of a sentence starting with a comma. German sentences often require turning around and in CAT that means looking about three segments (at best) further to go and find what the gist of the thing is exactly. That is irritating. Not only that, but your TM will contain faulty translations. And yes, you can start merging, but then I am faster doing it the other way. It should be 'Computer ASSISTED translation' and not me assisting the assisted translation...
Also, if the text is badly written, you end up having to merge sentences, but that means typing one thing in one segment and typing something else in the next. Annoying. If you are doing German you are probably going to have to split sentences, having the same trouble.
The problem is that CAT tools seem a good idea on paper, thought out by monolingual people, but in practice they do not work for the simple reason that languages are not exactly the same in punctuation nor even sentence building and how much information there is in a sentence at all.
We for two (:D) produce bad texts if we work in CAT and we are not careful. ▲ Collapse
| | |
|
|
|
Alex Lago
Spain
Local time: 11:08
English to Spanish
+ ...
| Tools do that | Jan 26, 2012 |
This is a collateral effect of a lot of tools.
Take for example mobile telephones, those of us who worked in the days before mobile telephones remember the days when you knew everybody's telephone number (anyone you called on a regular basis), now with the phone numbers stored in my mobile there are very few numbers I know by heart, I just look for someone's name and dial but if I had to tell you their number I wouldn't have a clue, even for people who I call on a daily basis.
| | | |
| How much different is it really working with a CAT? | Jan 26, 2012 |
Hello,
I have only ever worked with Wordfast Classic, which is very close to working with a Word document, so I don't know about other tools.
I don't think that this is all that much different from the work that I used to do without a CAT. With or without a CAT, you read through the entire text before starting and may make some notes in the text or on paper. (With a CAT, you have to also make sure that there are no superfluous paragraph breaks in the source text.) After... See more Hello,
I have only ever worked with Wordfast Classic, which is very close to working with a Word document, so I don't know about other tools.
I don't think that this is all that much different from the work that I used to do without a CAT. With or without a CAT, you read through the entire text before starting and may make some notes in the text or on paper. (With a CAT, you have to also make sure that there are no superfluous paragraph breaks in the source text.) After that, most work is focused at the level of the sentence or on small groups of sentences or a single sentence broken into two or three new sentences. Any of this is no more difficult to organize with a CAT than typing directing into an original text. Are other translators really consistently moving whole sentences or bits of information more than two or three sentences up or down in the translation?
Afterwards, the target text can be viewed as an integral whole, with the source text hidden. Here, you can check for macrotextual issues like repeated words, transitions, etc. - if you need to make a change or check the source, you can toggle back to the combined source-and-target view. After that, you can clean up the file and do a final spellcheck and formatting check.
If the job involves significant substantive or organizational editing, then this needs to be done either in the source text before segmenting or after the text has been cleaned up, with the target-language text. Both of these lead to problems with the TMs, but I think that most of the people who are against CATs work with texts where TMs are of little use anyway. And once again, this process is not that much different from working directly in a source document - anyone who carried out this kind of changes in the middle of the translation would invariably lose text every now and then.
The advantages of CATs are that they make it very hard to forget source-text in the translation, very easy to check target- vs. source-text, very easy to use a glossary and a concordance to check how you translated something three years ago (CATs are indeed bad for your memory, but they're good for your heartrate), very easy to avoid typos while transferring source into target text, and they are also a big help if a text that has already been translated is revised or is cannibalized in a later text by the same author.
CATs might reinforce some bad habits; but, if you invest a little time in a good CAT, it won't teach you any new bad habits.
Sincerely,
Michael ▲ Collapse
| | | |
S P Willcock (X)
Local time: 12:08
German to English
+ ...
| monolingual mindset | Jan 26, 2012 |
Kirsten Bodart wrote:
The problem is that CAT tools seem a good idea on paper, thought out by monolingual people, but in practice they do not work for the simple reason that languages are not exactly the same in punctuation nor even sentence building and how much information there is in a sentence at all. Absolutely, especially your point about the monolingual mindset of the designers, who may perhaps have learned further langauges but don't seem to savour them. Which is why, to return to my earlier point, the software houses don't realise that releasing a training video with a presenter who doesn''t speak the language she's "working" in shows indifference to the pont of contempt towards us, the users.
It's like employing a shoe saleswoman who knows all about knotting laces but has never actually walked anywhere.
| | | |
| Pages in topic: [1 2] > |
Login to reply/comment 
Translation news
