| Pages in topic: < [1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183] > | Should “native language” claims be verified? Thread poster: XXXphxxx (X)
|
|---|
Balasubramaniam L. 
India
Local time: 05:10
Member (2006)
English to Hindi
+ ...
SITE LOCALIZER | We will have to drop the idea of native language, of course... | Sep 11, 2012 |
Phil Hand wrote:
The problem is this: a system like that would be likely to become in practice the standard for nativeness. Once people realised that all you have to do to be allowed to call yourself native is to pass the grammar test, there would be no incentive to be honest. Anyone who passes would almost certainly put the languages they pass in up as native.
I like Samuel's idea too. As always, it creatively tackles the problem without becoming ideological or dogmatic.
The answer to Phil's above reservation to it is that we will of course then have to drop using native language for translator selection.
If we stop using native language in this way, then no one who passes this test would see any advantage in claiming that the language whose test they have passed is their native language. That is the crucial requirement in my view for implementing this idea - we move away from native language and concentrate on demonstrable ways of testing language competency. If translators still want to list native language, they can do that in their profile. But it will not be verified by the site, nor will it be used in any way to control access to jobs.
I also like the idea of repeating the test every year. A variation of this could be to allow outsourcers to use it for every job posting. That is, outsourcers can ask job bidders to take this test as a condition of applying for the job and they can evaluate the scores of the test while deciding on their translator.
On the practical side, this idea would be difficult to implement because of the effort and expertise that would be required to put together the question bank for each language. There are I think more than 200 languages listed in proz.com. Preparing question banks of 5,000+ questions for each of them won't be a trivial task.
Maybe, we can limit this idea to the more problematic languages such as English, Chinese, etc.
| | | | | verification of native language, not translation competence or use of basic grammar | Sep 11, 2012 |
Samuel Murray wrote:
What is a problem in some languages is not a problem in other languages, and consequently the solutions that apply to some languages do not apply to other languages. This was perhaps one value of this thread, to show and share how although most languages have a problem with native language declaration, they don't all have the same problem with it.
Samuel
The problem we established at the beginning of this thread is that people are intentionally lying about their native language. Not more, not less.
I get the feeling you are definitely not for verifying native languages.
Samuel Murray wrote:
I have met a few translators in my own native language whose nativeness (according to the linguistic definitions that you so dearly love) is not in question, and who can speak their native language fluently, but whose written language is so bad that I would not have been surprised to be told that their translations have to be redone from scratch every time. The proposed test in this sub-thread would weed out such translators, whether they are natives or not. The only possible relevance that this has for the native language issue is that many such translators who fraudulantly claim to be natives in that language would be weeded out also.
That is all beside the point (IMHO). The criterion that is sought and that I as a native speaker hold important is indeed the "native language" criterion. You can try to discredit it by saying that native language cannot guarantee translation quality which you clearly do in your statement above. This is the same old misunderstanding again.
The reason those bozos above didn't produce good translations was that they were not professional translators. But what the client might have specified was a) native speaker and b) expertise in marketing.
You won't be able to change this wishlist. You can try to discredit native language because some natives screwed up badly but that doesn't make the criterion itself unimportant.
Native language is a valuable criterion as long as the other important factors such as experience and expertise are in place as well. That should be a given for professional translators.
I don't know what else to say to convince you that native language as the only criterion for choosing a translator is a complete misinterpretation of the native language criterion. It is always to be understood in combination with the other required translator criteria.
A professional translator with experience and expertise uses his/her native language competence as an additional valuable asset. And the translation market understands that too.
It means nothing if it refers to a non-translator, or a fake translator who is either a native speaker or a non-native lying about his/her native language.
But it is absolutely important if it is part of a professional translator's portfolio.
Thus, by claiming it here on a translator portal, non-natives who pretend to be natives imply not "just" that they are native speakers, no, they're implying that their special native language competence tops all their other great assets, such as experience and expertise because they must be translators, why else would they be here.
Trust me, professionals are not going to take a grammar test (why should they?) and they're not going to simply let go of the so "unimportant" native language credential or selection criterion.
Proz.com said that in the future, native language claims of two languages must be confirmed before native language peers. We're waiting for it. We will certainly not wait forever. I think Proz.com might understand that.
B
[Edited at 2012-09-11 04:31 GMT]
| | | | Balasubramaniam L.
India
Local time: 05:10
Member (2006)
English to Hindi
+ ...
SITE LOCALIZER | Translation as a product vs translation as an artefact | Sep 11, 2012 |
As some are still not convinced sufficiently as to why in the modern world of translation native language skill is losing its shine, here is another argument to consider.
Consider a cobbler making a shoe and let us move back in time when this was the only option available for making footwear. The more skilled the cobbler is, the better shoe you can expect to get from him. You can compare the cobbler’s skill with native language expertise. The cobbler intuitively understands the pr... See more As some are still not convinced sufficiently as to why in the modern world of translation native language skill is losing its shine, here is another argument to consider.
Consider a cobbler making a shoe and let us move back in time when this was the only option available for making footwear. The more skilled the cobbler is, the better shoe you can expect to get from him. You can compare the cobbler’s skill with native language expertise. The cobbler intuitively understands the process of shoe-making, his family has been doing it for generations and they have deep insights into the process which no upstart cobbler (read non-native translator) can ever understand, let alone replicate.
Now consider the era after Henry Ford has implemented the assembly line process of manufacturing. In this type of manufacture, the product that is manufactured is not the work of the hands or skills of a single workman. There are the designer, the engineer, the metallurgist, the software programmer, the manager, the financier, and a host of other people, whose skills go into the production of a satisfactory car.
Let us consider the specific case of a translation product. It can be done in two ways. One is the cobbler way, where the finished product is the result of the work of one workman, the translator. And there is the Ford way, in which the translation is done by one person, it is then reviewed by a subject reviewer, the language is polished up by a copy-editor and finally it is proofread by a proof-reader. The end result of this process is also as good a translation as produced by the cobbler-translator.
Also, while all types of translation cannot be tackled by the cobbler-translator, all types of translation can be tackled by the Ford method as in this method, no skill-set is a limiting factor in producing a finished, satisfactory translation. Needless to say, most translations in the modern translation industry are done by the latter method and artisan translation where native language skills can be expected to be critical are becoming more and more a rarity.
The same thing has happened in many comparable industries – software for example. When Bill Gates started out, he was an artisan programmer who sat in his garage and developed the first version of the DOS operating system. Software development is no longer done in this fashion. It is a collaborative effort between several experts who together overcome the shortage of any one particular skill.
This is why outsourcers don’t give much importance to native language skills. There are effective ways to circumvent lack of native language skills.
[2012-09-11 05:44 GMT पर संपादन हुआ] ▲ Collapse
| | | | Balasubramaniam L.
India
Local time: 05:10
Member (2006)
English to Hindi
+ ...
SITE LOCALIZER | And there will soon be machine translation to consider | Sep 11, 2012 |
Eventually machine translation is going to get sufficiently good to be able to produce the first version of all but the most demanding type of translation. This first version will then be edited by human hands to produce the final version.
This is comparable to the current situation where translations are done by non-native hands which are then vetted by language experts.
Are we then going to try to stem the march of progress and say that machine translation is not to b... See more Eventually machine translation is going to get sufficiently good to be able to produce the first version of all but the most demanding type of translation. This first version will then be edited by human hands to produce the final version.
This is comparable to the current situation where translations are done by non-native hands which are then vetted by language experts.
Are we then going to try to stem the march of progress and say that machine translation is not to be allowed because it is not done by native translators (or should I say native machines!)? Clearly it is pointless to take such a stance.
We will have to take it as a given in our profession that there is going to be continuous improvements and changes in the translation process and we cannot bank on exclusionary skills like native language.
[2012-09-11 05:12 GMT पर संपादन हुआ] ▲ Collapse
| | |
|
|
|
Samuel Murray
Netherlands
Local time: 01:40
Member (2006)
English to Afrikaans
+ ...
Charlie Bavington wrote: Samuel Murray wrote:
Allow all translators to create profiles and use the site features, but disallow translators from bidding or being found in the search results if they haven't passed a basic language skill test. People who have already paid membership of the site ... are probably gonna take a pretty dim view of either having part of that service withdrawn or having to jump through extra hoops to get that service, all because other people are misrepresenting themselves.
So your objection basically means that the test should only apply to new people as well as people whose membership runs out. Not really a big objection.
But even so, I don't share your view that people would "probably" object to this test. Here's why:
* If you can truly translate in a language, you would ace the test. So there is practically no effort involved (just an hour of your time, and let's give people a full month to complete the test before the sanction applies).
* Everyone benefits (except those being weeded out) if ProZ.com weeds out bad translators, because it improves ProZ.com's reputation in the medium and long term, leading to more clients using the site, leading to more jobs. This test isn't something that you must give, but something that you GET for your membership fee.
* Another benefit for non-weeded users is that the number of users bidding on jobs or being displayed to clients who search for translators is reduced.
* If you want, you could exempt red pee members from the first test or some of the tests.
I took a couple of lines from this file to show what three questions in the English test might be (some of those who agree with me here might have had different question in mind -- please comment if you will):
1. Indicate the sentence(s) in which the singular verb is used incorrectly:
a A number of measurements has been made.
b It appears that 30 g is required daily.
c The most part of the compounds is inactive.
2. Indicate the sentence(s) in which the singular verb is used incorrectly:
a The majority of the tests was performed in London.
b In some areas 60% of the workforce is unemployed.
c Acoustics is the study of the behaviour of sound.
3. Indicate the sentence(s) in which the singular verb is used incorrectly:
a The number of units has been reported to be 50.
b The acoustics in this hall is terrible.
c A total of 32 measurements has been made.
It would be interesting if English natives in this thread can tell us how long it took them to answer this, and what their answers would be. And also which of these options you would have deemed unfair, i.e. either singular or plural verb would have been correct. The idea is that such a test would contain no trick questions.
...you're still likely to be administrating thousands or tests.
What part of "automated" don't you understand?
Once the tests have been written each year (and stress tested by some volunteers to weed out ambiguous questions), the tests would practically run themselves, on the web site. You'd visit a web page that contains the test, most of the questions would be multiple option question, some might include typing one or two words to complete a given sentence, and the answers can be tallied within seconds after the test is completed.
Samuel
| | | | Samuel Murray
Netherlands
Local time: 01:40
Member (2006)
English to Afrikaans
+ ...
Charlie Bavington wrote:
...but in essence, agreement to enforce the "no misrepresntation" rule has to come from the top first, since confirming others' representations is what it all boils down to, does it not?
How is what you're proposing different from rule #1 in the ProZ.com user agreement? See here, it says clearly (and it says it as the very first rule) "You may not misrepresent yourself or your firm."
Or do you mean that ProZ.com should be more vocal about it?
Just saying that ProZ.com must more strictly enforce it is rather fuzzy, and not much of a solution, since it doesn't make it easier to sanction misrepresentation than it currently is. Those things that are difficult to measure will still be just as difficult to measure, no matter how firmly the site commits itself to stricter enforcement.
In fact, some things are so difficult to verify that it is an impossible task to police it. So instead of trying to verify the unverifiable, why not verify something verifiable that is a natural prerequisite for the thing that is unverifiable? That is why I propose the test as a small extra step that will be easy to administer and which would not disfavour true translators in any way.
Samuel
| | | | Samuel Murray
Netherlands
Local time: 01:40
Member (2006)
English to Afrikaans
+ ...
Bernhard Sulzer wrote: Samuel Murray wrote:
This was perhaps one value of this thread, to show and share how although most languages have a problem with native language declaration, they don't all have the same problem with it. The problem we established at the beginning of this thread is that people are intentionally lying about their native language. Not more, not less.
Yes, yes, I understand... for you the issue of native language is about ideology, not more, not less. The practical implications of the various aspects of native language is not of interest to you, as long as the sanctity of "native language" as a concept remains untouched. Do I understand your point of view correctly?
My reply to Michele would not suit you, obviously, since Michele's view on native language is different from yours -- she is not religious about it, although she does tend to have a more academic, theoretical view of the issues at hand than some of the other participants to whose point of view I have responded in the past day or two.
I get the feeling you are definitely not for verifying native languages.
The fact that I consider methods of solving native language related problems that do not involve verifying native language does not say anything about my opinion about verifying native language.
Samuel Murray wrote:
I have met a few translators in my own native language whose nativeness (according to the linguistic definitions that you so dearly love) is not in question, and who can speak their native language fluently, but whose written language is so bad that... You can try to discredit it by saying that native language cannot guarantee translation quality which you clearly do in your statement above.
I don't think admitting or stating that some native speakers are terrible writers of it is any way a discredit to the idea of native language. Do you?
I don't know what else to say to convince you that native language as the only criterion for choosing a translator is a complete misinterpretation of the native language criterion.
I don't know what to say to convince you that what you seem to think my opinion on this matter is is in fact nothing of the sort. I think you're confusing me with someone else.
Samuel
| | | | Ty Kendall
United Kingdom
Local time: 00:40
Hebrew to English
Balasubramaniam L. wrote:
You can compare the cobbler’s skill with native language expertise.
No, you can't.
The end result of this process is also as good a translation as produced by the cobbler-translator.
If it has to go through THAT many stages just to get it to the same level then you sure aren't getting value for money as it would surely cost a lot more. This is a disincentive, not an incentive for non-native translators.
Also, while all types of translation cannot be tackled by the cobbler-translator, all types of translation can be tackled by the Ford method as in this method
Says who?
Needless to say, most translations in the modern translation industry are done by the latter method and artisan translation where native language skills can be expected to be critical are becoming more and more a rarity
Says who? It is very much needed to say...show me your data to substantiate these claims.
This is why outsourcers don’t give much importance to native language skills.
But they do, I think that's been pretty solidly established and accepted by both sides of the argument.
There are effective ways to circumvent lack of native language skills.
Circumvent...but not compensate for? Interesting.
[Edited at 2012-09-11 06:58 GMT]
| | |
|
|
|
Samuel Murray
Netherlands
Local time: 01:40
Member (2006)
English to Afrikaans
+ ...
| @Balasubramaniam | Sep 11, 2012 |
Balasubramaniam L. wrote:
The answer to Phil's ... reservation to it is that we will of course then have to drop using native language for translator selection.
If we stop using native language in this way, then no one who passes this test would see any advantage in claiming that the language whose test they have passed is their native language.
No, I must object to this (see also my answer to Phil's reservation). I thank you for supporting my idea, but please note that my idea is not meant as a replacement of the native language option, and it is not meant to support it either. Native language is very important for both translators and clients, and giving it prominence (or at least a place) on the site is simply logical.
I also like the idea of repeating the test every year. A variation of this could be to allow outsourcers to use it for every job posting.
I actually considered that, but:
* This goes against the main idea of the test, namely that it would exclude such poor translators from the searches altogether.
* I also think a single test once a year would go down better than having to pass a silly test every time you bid on a job.
* Also, having a little test before each job would increase the number of appeals, which adds to administration.
* Also (and this is quite important, in fact), a weak candidate is far more likely to strike it lucky when doing a quick, short test than when doing a longer test.
* And a longer test is also fairer to borderline cases, because you can answer more questions incorrectly and still qualify, as long as your average score is high.
On the practical side, [there is] ... the effort and expertise that would be required to put together the question bank for each language. ... Maybe, we can limit this idea to the more problematic languages such as English, Chinese, etc.
Yes, writing the initial test would take a while. One can protect that effort by randomising the questions for each candidate, and by not showing the answers (just the final score). If a test has been stress tested by a large number of volunteers to ensure that there are no ambiguities and no regionalisms in it, it is not necessary to show failed candidates which questions they got wrong.
And yes, it makes sense to limit the tests to languages with an abundance of translators. And I also think that for some languages regional variants simply has to come into play (e.g. PT).
| | | | Phil Hand
China
Local time: 07:40
Chinese to English
| But I'm with Charlie on this one | Sep 11, 2012 |
I like your solution, Samuel, in that it does look like it would achieve what I've said I want to achieve all along. Plus it could be automated, so it would actually be pretty easy to run once it was set up.
However, I don't think it's practical, for a couple of reasons.
1) People hate tests. Never having to take exams again is one of the great joys of being a grown up. I think people will revolt - it doesn't matter how easy/hard the test is, just the very notion of a t... See more I like your solution, Samuel, in that it does look like it would achieve what I've said I want to achieve all along. Plus it could be automated, so it would actually be pretty easy to run once it was set up.
However, I don't think it's practical, for a couple of reasons.
1) People hate tests. Never having to take exams again is one of the great joys of being a grown up. I think people will revolt - it doesn't matter how easy/hard the test is, just the very notion of a test gets people's teeth on edge.
2) You would knock a lot of people (or a lot of their combinations) out of the database. The transition would be rough for Proz, and I don't think the site owners would want it.
So I would find this proposal difficult to support. I was taking it as a useful step forward in the debate, rather than an endpoint! ▲ Collapse
| | | | Michele Fauble
United States
Local time: 16:40
Member (2006)
Norwegian to English
+ ...
| Non-native error | Sep 11, 2012 |
Samuel Murray wrote:
It would be interesting if English natives in this thread can tell us how long it took them to answer this, and what their answers would be.
Glancing through the questions, I immediately spotted the non-native phrasing The most part of the compounds ... and being a translator of Scandinavian languages, I recognized its Scandinavian origins (mestaparten av ...). And the questions are indeed taken from a Swedish site.
| | | | Samuel Murray
Netherlands
Local time: 01:40
Member (2006)
English to Afrikaans
+ ...
| @Balasubramaniam, re the cobbler theory | Sep 11, 2012 |
Balasubramaniam L. wrote:
One is the cobbler way, where the finished product is the result of the work of one workman, the translator. And there is the Ford way, in which the translation is done by one person, it is then reviewed by a subject reviewer, the language is polished up by a copy-editor and finally it is proofread by a proof-reader.
This is indeed how some translation houses and even the translation divisions of large, successful multinational corporatoins operate: errors made by translators are simply fixed by the reviewers. There are real dangers to this approach, but it is quite a common approach nonetheless (far more common than some would like to admit).
Clients who use this type of system really only need third-rate translators. However, not all clients work this way, and such clients need to use first-rate translators. For mill clients who demand third-rate translators, nativeness is not an issue, because the QA compensates for it, but for the vast majority of clients, nativeness (whether in the translator or in the translation) it is still highly relevant.
The same thing has happened in many comparable industries – software for example. When Bill Gates started out, he was an artisan programmer who sat in his garage and developed the first version of the DOS operating system. Software development is no longer done in this fashion.
I'm afraid translation is more complex than that.
You can take an intelligent Indian who knows nothing of software programming and turn him into a capable programmer in just 1 year, capable of designing and troubleshooting software to the requirements of most clients.
Your argment would become valid when translator training matures to the point where... you can take an intelligent Indian who knows nothing of translation and who can't speak either German or Spanish, and turn him into a German-Spanish translator in just 1 year, capable of translating and proofreading texts to the requirements of most clients.
| | |
|
|
|
Samuel Murray
Netherlands
Local time: 01:40
Member (2006)
English to Afrikaans
+ ...
| @Michele -- I also thought that that was odd | Sep 11, 2012 |
Michele Fauble wrote:
Glancing through the questions, I immediately spotted the non-native phrasing The most part of the compounds ... and being a translator of Scandinavian languages, I recognized its Scandinavian origins (mestaparten av ...). And the questions are indeed taken from a Swedish site.
I also thought that that was an odd phrasing, and I was tempted to "fix" it, but in the end I just pasted the questions as-is. They are, after all, simply an illustration of the type of questions.
| | | | XXXphxxx (X)
United Kingdom
Local time: 00:40
Portuguese to English
+ ...
TOPIC STARTER | The Certified Pros | Sep 11, 2012 |
A couple of you have mentioned exempting the red Ps from these tests, some of whom (IMHO) would appear to be the worst offenders and have perhaps sought the badge as a way of affording them greater protection from being challenged. Please don’t look at the P scheme to see how things ought to be done. My definite impression over the last couple of months of this debate is that the proportion of Certified Pros misrepresenting themselves as... See more A couple of you have mentioned exempting the red Ps from these tests, some of whom (IMHO) would appear to be the worst offenders and have perhaps sought the badge as a way of affording them greater protection from being challenged. Please don’t look at the P scheme to see how things ought to be done. My definite impression over the last couple of months of this debate is that the proportion of Certified Pros misrepresenting themselves as regards their native language is no lower than amongst the rest of registered users. ▲ Collapse
| | | | XXXphxxx (X)
United Kingdom
Local time: 00:40
Portuguese to English
+ ...
TOPIC STARTER
Yaotl Altan wrote:
WE must vote and end this once and for all in order to know if we make changes or not to the profiles in Proz regarding Native languages.
Anyone interested in running a poll, as a potential prelude to a petition? Any consensus on the wording?
| | | | | Pages in topic: < [1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183] > | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » Should “native language” claims be verified? | Trados Studio 2022 Freelance | The leading translation software used by over 270,000 translators.
Designed with your feedback in mind, Trados Studio 2022 delivers an unrivalled, powerful desktop
and cloud solution, empowering you to work in the most efficient and cost-effective way.
More info » |
| | 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 
