| Pages in topic: < [1 2 3 4 5] > | Voting just for "quality of writing" is totally unacceptable Thread poster: Roland Nienerza
|
|---|
Caro maestro,
I appreciate that you took the trouble to look into this discussion.
The best thing about criticism is certainly when it comes with a solution attached.
In my last answer I had hit on the possibility that errant votes just on the basis of the nicest wrappings, regardless of what is inside the package, could be reduced if any vote on "quality of writing" had to be accompanied by a vote on "accuracy of translation".
As a matter of f... See more Caro maestro,
I appreciate that you took the trouble to look into this discussion.
The best thing about criticism is certainly when it comes with a solution attached.
In my last answer I had hit on the possibility that errant votes just on the basis of the nicest wrappings, regardless of what is inside the package, could be reduced if any vote on "quality of writing" had to be accompanied by a vote on "accuracy of translation".
As a matter of fact, the technical gizmo for that is already in the contest software.
All one would have to do is to interlock the already existing two five star rows, for quality and accuracy, in a way that no separate voting will be possible any more. - Thus, whoever wishes to vote on either, has to vote on the other as well.
That would need just a tiny programming input - for a considerable result. ▲ Collapse
| | | | | Just skip the category then | Dec 10, 2008 |
Roland Nienerza wrote:
In my last answer I had hit on the possibility that errant votes just on the basis of the nicest wrappings, regardless of what is inside the package, could be reduced if any vote on "quality of writing" had to be accompanied by a vote on "accuracy of translation".
I'm sorry, I still don't get it... are you suggesting that one's vote on "quality of writing" should be influenced by one's estimation of the accuracy of the translation? A *very* nicely written text (well flowing, no grammar mistakes, no spelling mistakes, no typos, no punctuation errors) that happens to be an inadequate translation should be rated as "bad writing quality"? What's the point of that category then? It'd just be another "quality of translation" rating. Why not just skip the quality of writing and go for 100% quality of translation? Or "quality of final product", if you want to make it a bit more inclusive of such things as grammar and spelling. Would be fine with me. I just still don't see the point of interlocking the two categories when they have absolutely nothing to do with each other. They result in an average score each, and as far as I understand it those scores are not then added up to make one single final score. Or are they?
I would think that a person with an average rating of five stars on quality of writing and an average of one in quality of translation would sorely lose against a contestant whose ratings are the other way around. At least, if we assume that those two are the only contestants. Am I right, or am I just not getting how the rating system works?
| | | |
Annika Persson wrote: Roland Nienerza wrote:
In my last answer I had hit on the possibility that errant votes just on the basis of the nicest wrappings, regardless of what is inside the package, could be reduced if any vote on "quality of writing" had to be accompanied by a vote on "accuracy of translation".
I'm sorry, I still don't get it... are you suggesting that one's vote on "quality of writing" should be influenced by one's estimation of the accuracy of the translation?
I would not say "influenced" - but accompanied. - I think that there should not be a vote on "quality of writing" without a vote on "accurateness of translation".
Annika Persson wrote:
A *very* nicely written text (well flowing, no grammar mistakes, no spelling mistakes, no typos, no punctuation errors) that happens to be an inadequate translation should be rated as "bad writing quality"?
Not for me. -
"A *very* nicely written text (well flowing, no grammar mistakes, no spelling mistakes, no typos, no punctuation errors) that happens to be an inadequate translation should be rated as a " *very* nicely written text" + an "inadequate translation".
Annika Persson wrote:
What's the point of that category then? It'd just be another "quality of translation" rating. Why not just skip the quality of writing and go for 100% quality of translation? Or "quality of final product", if you want to make it a bit more inclusive of such things as grammar and spelling. Would be fine with me.
It would indeed with me. - At least a differentiated voting should be summed up in a comprehensive score.
[quote]Annika Persson wrote:
I just still don't see the point of interlocking the two categories when they have absolutely nothing to do with each other. They result in an average score each, and as far as I understand it those scores are not then added up to make one single final score. Or are they?[quote]
This indeed I have not understood yet either. - The interlocking idea was just for preventing that someone votes positively just on the surface look of the translation. They should alway vote for both. - But those who do not know the source language cannot do that.
Annika Persson wrote:
I would think that a person with an average rating of five stars on quality of writing and an average of one in quality of translation would sorely lose against a contestant whose ratings are the other way around. At least, if we assume that those two are the only contestants. Am I right, or am I just not getting how the rating system works?
Well, I would think that a person "with an average rating of five stars on quality of writing and an average of one in quality of translation" would not sorely ***but very justly and deservedly*** lose against a contestant with votings the other way round.
As said, how "quality of writing" and "accurateness of translation" are weighted I did not grasp yet either. - From the qualification I understood that the final score is established by computing the average for each separately, adding the two averages and dividing them by two. - If I understand this correctly, "quality of writing" and "accruateness of translation" are equally rated.
| | | | | Translation should weigh heavier | Dec 10, 2008 |
Roland Nienerza wrote:
As said, how "quality of writing" and "accurateness of translation" are weighted I did not grasp yet either. - From the qualification I understood that the final score is established by computing the average for each separately, adding the two averages and dividing them by two. - If I understand this correctly, "quality of writing" and "accruateness of translation" are equally rated.
Well, if that is the case, then of course I at least agree on that point: quality of translation should weigh heavier than quality of writing.
However, I still think that it is totally fine to just rate the quality of writing regardless of the source text. As long as the quality of translation (as a category of its own) weighs heavier, that is. The more votes, the more accurate the average results, and that can only be positive (makes it less likely that someone earns more stars than they deserve). It is a pity that the amount of competent voters on quality of translation cannot be raised by such simple means as well, but that's the way it is. The accuracy of the average on quality of translation will not get better by excluding those who are "only" native speakers from rating the quality of writing. It is bad enough that one category must suffer from a potentially low number of voters - but why drag down the other one together with the first?
| | |
|
|
|
| No reason to link the categories | Dec 10, 2008 |
I would give equal weight to the two areas. An "accurate" translation that sounds like garbage will often not be trusted and fail in its purpose in many instances. Certainly for literature or articles that will be the case. I really doubt that the intention of the authors in these areas is to produce work that reads badly - if one ignores certain experimental forms of "literature" like random cutting and pasting.
There is also no reason that I see for demanding that a vote be cast i... See more I would give equal weight to the two areas. An "accurate" translation that sounds like garbage will often not be trusted and fail in its purpose in many instances. Certainly for literature or articles that will be the case. I really doubt that the intention of the authors in these areas is to produce work that reads badly - if one ignores certain experimental forms of "literature" like random cutting and pasting.
There is also no reason that I see for demanding that a vote be cast in both categories. There are probably enough people in some cases who think they are able to understand a source text and make a judgment who actually are not that I don't think you are going to achieve an ideal configuration of voters with any scenario. I think the idea of letting persons competent in the language of the target text judge the translation as a piece of work on its own without reference to the source text is a marvelous idea. That's how our work will be judged in the real world most of the time - by people who are not competent in the source language - so let's GET REAL and drop the petty academic foolishness. ▲ Collapse
| | | | | "GET REAL" - progressing from "nonsense" over "absurdity" to "petty academic foolishness" - | Dec 10, 2008 |
Kevin Lossner wrote:
I think the idea of letting persons competent in the language of the target text judge the translation as a piece of work on its own without reference to the source text is a marvelous idea. That's how our work will be judged in the real world most of the time - by people who are not competent in the source language - so let's GET REAL and drop the petty academic foolishness.
Letting "persons competent in the language of the target text judge the translation as a **piece of work on its own** without reference to the source is a marvelous idea".
This is indeed a marvelous idea. - And a very funny and pathetic one to that.
As long as indeed they judge it as **piece of work on its own**, whatever that may be - ha, ha, and not as a translation.
For that would mean to "GET REAL" cheating, and not just a "petty academic foolishness".
[Edited at 2008-12-11 14:37 GMT]
| | | | Kristina Kolic 
Croatia
Local time: 14:39
English to Croatian
+ ...
SITE LOCALIZER | "Quality of writing" involves tagging likes and dislikes | Dec 11, 2008 |
Roland Nienerza wrote:
I think that there should not be a vote on "quality of writing" without a vote on "accurateness of translation". (...)
A *very* nicely written text (well flowing, no grammar mistakes, no spelling mistakes, no typos, no punctuation errors) that happens to be an inadequate translation should be rated as a " *very* nicely written text" + an "inadequate translation".
I am not sure whether you actually had the opportunity to participate in voting on "quality of writing", but it seems to me that you are playing with words without knowing the actual process involved in "quality of writing"...
I had the opportunity to "vote" on "quality of writing" in a language pair and I actually did use the like and dislike tags not only for grammar, spelling or punctuation errors but also for inadequate translations. So, even if my votes in this "quality of writing" count less, I did point to mistranslations or inadequate translations and they certainly did influence my vote.
| | | | | Thanks, Roland | Dec 11, 2008 |
Roland Nienerza wrote:
In my last answer I had hit on the possibility that errant votes just on the basis of the nicest wrappings, regardless of what is inside the package, could be reduced if any vote on "quality of writing" had to be accompanied by a vote on "accuracy of translation".
Thank you for your attention to this issue.
On questions like this the thing to do is to look at empirical data. If we had a substantial number of ratings from people on entries into their target languages, I would encourage the two current members of our contest team to observe the desirability of this form of voting.
As it stands, with the interface being somewhat counter-intuitive, a relatively small number of people seem to have taken up the opportunity to rate "quality of writing" only in their non-working pairs. So we don't have much data to look at.
I suppose this is good news, if one thinks that such voting would be bad for the contest. In that case, however, there is also some bad news: I have asked the team to clarify the interface.
Given the level of attention and the points made on this issue (and anyway), I can assure you we'll be looking closely in the future at whether or not this form of rating is a good thing. If it turns out to be biasing outcomes in an undesirable way, we'll do away with it (perhaps using your suggestion.)
Thanks, Roland, and everyone who participated in this discussion!
| | |
|
|
|
| Always a good approach | Dec 11, 2008 |
Henry D wrote:
On questions like this the thing to do is to look at empirical data.
So the plan is to look at what's really going on and act accordingly as opposed to following unsubstantiated assumptions however plausible? Sounds like a good idea.
| | | | Sara Noss
United Kingdom
Local time: 13:39
Member (2006)
French to English
+ ...
| Accuracy and quality of writing, | Dec 11, 2008 |
like Fred and Ginger, gin and tonic and a plethora of other more or less famous and unbeatable pairings, produce their best results when they go together. The extent to which this consideration is employed by those voting in the contests will always be a source of debate on ProZ.com and rightly so, in some cases. To rate a translation without any reference to the source text, which should be well known to me, would not occur to me, as it would not occur to me to ignore either accuracy or quality... See more like Fred and Ginger, gin and tonic and a plethora of other more or less famous and unbeatable pairings, produce their best results when they go together. The extent to which this consideration is employed by those voting in the contests will always be a source of debate on ProZ.com and rightly so, in some cases. To rate a translation without any reference to the source text, which should be well known to me, would not occur to me, as it would not occur to me to ignore either accuracy or quality. I think this must be (has to be) the approach of many a translator. When looking at the end result of a translation, I appreciate not only the translator's close understanding of a text, but also their ability to do justice to the author and their target audience and this can, and often does, necessitate some editing. However, this can and must be done with a judicious and sensitive hand and without creating your own particular work of fiction.
Looking into the empirical data, and getting a clear idea of what really is happening with community voting patterns is a great idea not only to enhance the contest, but possibly as a means for all of us to reassure ourselves that our voting instincts are rather less random than we might expect.
Kind regards,
Sara
[Edited at 2008-12-11 14:32 GMT] ▲ Collapse
| | | | Heinrich Pesch
Finland
Local time: 15:39
Member (2003)
Finnish to German
+ ...
| How about the real thing? | Dec 11, 2008 |
When a publisher publishes the translation of a work of fiction, does s/he first read the original? No way, if the source language is something else than English I would be surprised if anyone at the publisher's is able to read the original at all.
So we are at the mercy of the translator.
But at this contest, how probable is it, that someone would deliver a very free adaptation? S/he would be found out immediately, sticking out between the rest. So this voting purely accordin... See more When a publisher publishes the translation of a work of fiction, does s/he first read the original? No way, if the source language is something else than English I would be surprised if anyone at the publisher's is able to read the original at all.
So we are at the mercy of the translator.
But at this contest, how probable is it, that someone would deliver a very free adaptation? S/he would be found out immediately, sticking out between the rest. So this voting purely according to quality of writing should work all right.
Regards
Heinrich ▲ Collapse
| | | | Caroline Mackay-Sim (X)
Local time: 22:39
English to French
+ ...
| Well said, Sara | Dec 11, 2008 |
Sara Noss wrote:
like Fred and Ginger, gin and tonic and a plethora of other more or less famous and unbeatable pairings, produce their best results when they go together. The extent to which this consideration is employed by those voting in the contests will always be a source of debate on ProZ.com and rightly so, in some cases. To rate a translation without any reference to the source text, which should be well known to me, would not occur to me, as it would not occur to me to ignore either accuracy or quality. I think this must be (has to be) the approach of many a translator. When looking at the end result of a translation, I appreciate not only the translator's close understanding of a text, but also their ability to do justice to the author and their target audience and this can, and often does, necessitate some editing. However, this can and must be done with a judicious and sensitive hand and without creating your own particular work of fiction.
Looking into the empirical data, and getting a clear idea of what really is happening with community voting patterns is a great idea not only to enhance the contest, but possibly as a means for all of us to reassure ourselves that our voting instincts are rather less random than we might expect.
Kind regards,
Sara
[Edited at 2008-12-11 14:32 GMT]
| | |
|
|
|
| If you really want to worry about accuracy, cheating, etc. | Dec 11, 2008 |
Consider that someone like me could probably declare Russian and Japanese (or any other languages for that matter) as working languages and synthesize a "translation" based on common elements of the target texts. A good writer could walk away with the honors barely being able to understand the source (or understanding it not at all!).
Not a scenario I would worry much about, even if it is possible. At some point, preferably an early one, I'll simply trust my colleagues to do the rig... See more Consider that someone like me could probably declare Russian and Japanese (or any other languages for that matter) as working languages and synthesize a "translation" based on common elements of the target texts. A good writer could walk away with the honors barely being able to understand the source (or understanding it not at all!).
Not a scenario I would worry much about, even if it is possible. At some point, preferably an early one, I'll simply trust my colleagues to do the right thing. As I see it, the real point of these contests is to have a bit of fun. If someone cuts me out by dirty deeds, I usually see the loss more as his.
In general I agree with Mr. Pesch on this one. I think the general ridicule one would receive from overstepping a clear boundary (if such a thing exists) would outweigh any benefit from flowery bullshit unrelated to a source text that takes the top prize. But as I noted before, I think considerable freedom in localization is legitimate, even if it might give the honored initiator of this thread a stroke. ▲ Collapse
| | | | | Accuracy and quality in translation should not be separated - | Dec 11, 2008 |
Dear cara,
Sara Noss wrote:
like Fred and Ginger, gin and tonic and a plethora of other more or less famous and unbeatable pairings, produce their best results when they go together.
Sure. - Two other really famous pairings of that kind that spring to the mind are Ginger and Ale as well as Johnny and Walker.;-)
But, as I said above, there are difference between these and translations.
While the above perform well in pairs, but could be and have been good even when acting on their own, a translation ultimately is not accurate, if badly written - while it could well be nicely written, without being accurate at all.
The value or even the beauty of a *piece of fiction* is in -
invention and imagination + quality of writing.
The value or even the beauty of a ***translation*** is in -
accuracy and truthfulness + quality of writing.
Do you see the difference?
Looking for the value of a *piece of fiction* you can take it as it stands.
Looking for the value of ***translation*** you ***have*** to look for the source text, or else you treat it as a *piece of fiction*;-).
And yet a word about accuracy. - Who says, that accuracy has to be destitute of "quality of writing"?
Accuracy, certainly in literary or other forms of non technical translation, resides in the elegance of precision, the dazzling beauty of rendering the thought or the concept of the source perfectly "one to one" - nothing less, but also nothing more. But that cannot be achieved with bumping, hobbled writing.
See what I mean-s?
And, as I said before. - There is a pretty reliable way of measuring the accuracy of a translation.
A way that quite obviously and surprisingly seems to be unknown by many, if not the great majority in the trade.
Baghs 'n baras -
R N
[Edited at 2008-12-11 16:05 GMT]
| | | | | If one really wants to worry about accuracy, cheating, etc. | Dec 11, 2008 |
Kevin Lossner wrote:
If someone cuts me out by dirty deeds, I usually see the loss more as his.
It need not really be "dirty deeds". - It is just "bad translations" from a professional, or for that matter, academical viewpoint. - Curiously enough, in this there might well be cheaters without really bad intentions. Merely a lack of proficiency.
Kevin Lossner wrote:
I think considerable freedom in localization is legitimate
If you made it not "considerable" but *necessary* or *adequate*, it would be alright for me.
Kevin Lossner wrote:
even if it might give the honored initiator of this thread a stroke.
It hardly would - even as he does not like certain strong words.
But it might cause him *crawl more than just one louse over the liver* - at times.
[Edited at 2008-12-11 16:08 GMT]
| | | | | Pages in topic: < [1 2 3 4 5] > | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » Voting just for "quality of writing" is totally unacceptable | CafeTran Espresso | You've never met a CAT tool this clever!
Translate faster & easier, using a sophisticated CAT tool built by a translator / developer.
Accept jobs from clients who use Trados, MemoQ, Wordfast & major CAT tools.
Download and start using CafeTran Espresso -- for free
Buy now! » |
| | TM-Town | Manage your TMs and Terms ... and boost your translation business
Are you ready for something fresh in the industry? TM-Town is a unique new site for you -- the freelance translator -- to store, manage and share translation memories (TMs) and glossaries...and potentially meet new clients on the basis of your prior work.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
