| 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)
|
|---|
Sheila Wilson 
Spain
Local time: 04:03
Member (2007)
English
+ ...
| A written chat would be best | Jul 3, 2012 |
Samuel Murray wrote:
The only way to make the written test as good as the spoken test would be to (a) first determine the translator's typing speed and (b) then giving him a strict time limit in which to do write the test. If you allow the translator 24 hours to do his written test, then it becomes nearly useless, since most translators would have been able to fix their errors by then.
I don't think that would prove a thing apart from perhaps that the deeper the pocket, the more native languages you could have.
Let each person who wants to be verified be interviewed for 10-15 minutes by 10 randomly selected translators who indicated that they are willing to help verify people. The test can consist of a bit of free chatting, a chat about current afairs, open-ended questions aimed at generating sufficient verbal content to judge the person's fluency. Lack of good vocabulary may be penalised but should not be the overall criteria of "fluency".
I think that would be the best test, although I would much prefer it to be in writing. As I said many pages ago, the differences between a high-level non-native speaker and a native speaker are sometimes difficult to spot when chatting but the non-nativeness becomes all too evident in writing. Most of us use instant messaging, don't we? I think true native speakers would be able to see beyond the odd typo. A typo is a typo - it doesn't alter your categorisation.
However, I think that ProZ still needs to tighten up the recommendations for completing this field, to cut down on the numbers being reviewed and the number of laughable claims.
Sheila
| | | | XXXphxxx (X)
United Kingdom
Local time: 04:03
Portuguese to English
+ ...
TOPIC STARTER
Robert Forstag wrote:
I don't see such a scheme as working for the following reasons:
1.
The existence of such buttons attached to the postings of those with one or more unverified native languages would seem to perpetuate an atmosphere of hostility and suspicion throughout the site, and even (it might be argued) imply a presumption of misrepresentation.
Agree.
2.
It begs the question of what exactly the criteria would be for verification/denial. Would 80% percent (for example) of affirmative responses to the multiple choice items based on a sample of 300 respondents be enough to constitute verification?
No idea. A site policy question I can't answer. I presume that's a support request.
3.
It would seem possible for a given individual desirous of verification to simply recruit his or her friends to provide positive input into such a system, and there would be no way of knowing if this has happened or not.
Agree, although one has to presume that the numbers going to such efforts and engaging in such high levels of deceit would be relatively small.
4.
Such a process provides no evaluation whatsoever of speech or listening comprehension (two critical language skills that ought to be assessed as part of any verification of a claim of native language ability).
Agree, although for the purposes of this site and their role as a translator I think it's their written skills we're interested in seeing and I still believe we can spot most non-natives at a hundred paces.
5.
Because such a scheme would evaluate only written content whose provenance could not be guaranteed, it would leave open the possibility that one seeking verification of a native language uploading a number of forum posts that have been edited by skilled native speakers in order to serve as (bogus) evidence of native skills, at least until such time that the desired verification has been obtained.
See point 3 above, i.e. possible, but not everyone is going to bother doing this.
In short, I don't see this kind of process as having much reliability or validity. I'm afraid this is a case in which, if it is to be done right, the validation system will need to go beyond computer-processed responses and actually involve direct human interaction and judgment calls.
I hear what you're saying and of course agree, that is the "ideal site" we spoke of before However, I think we have to be realistic and work within the limitations of what is possible in the near future and this Peer Review process is possibly the best we're likely to get for now.
Edited for HTML tags. Hopefully okay now.
[Edited at 2012-07-03 20:38 GMT]
[Edited at 2012-07-03 20:41 GMT]
| | | | Samuel Murray
Netherlands
Local time: 05:03
Member (2006)
English to Afrikaans
+ ...
Nicole Schnell wrote:
I was certain that you will get the hint immediately (huge compliment for being fast, as usual..).
Actually, I rarely read threads of more than 5 pages, and I rarely post more than 5 times unless I'm the thread starter, so you're actually quite lucky that I saw your reference as quickly as I had. Fortunately this thread has been very interesting (as you had said, too) and the people I disagree with here are intelligent and mostly eager to stay on-topic, so it has been a pleasure.
Then how come that self-declared native speaker issues are dissected to excess ... whereas truly brutal lies such as made-up years and decades of translation experience are perfectly fine and accepted?
There is no inconsistency. My approach to the part-time translator who has been a full-time nurse for 30 years is the same as my approach to the thousands upon thousands of translators who claim to have more than one native language: in the beginning, I assume good faith, until evidence shows otherwise. That is why I have hammered on definition for so long, and have rejected the advances of those who believe that anyone with a different opinion or interpretation of "native language" has got to be a liar and a cheat.
It makes perfect sense to me that most translators here are honest. Those who are out to make a quick buck often don't even fill in their profiles properly.
By the way, the "years of experience" issue has been discussed in the past, and I can remember from them that there was no consensus about how to declare that either. If you had been a part-time translator for 30 years, should you declare "15 years of experience" or "30 years of experience"? If you had a 3-year hiatus, do you subtract that from your "years of experience" or not (even if those three years were spent getting a translation degree)? If you had worked not 8 hours but 12 hours a day for the past two years, can you count that as 3 years of experience (and if not, why not)? That is another topic for discussion, but I acknowledge that mostly all sides have valid-sounding arguments and that there is not a single correct answer.
==
Nicole Schnell wrote:
If you want to be certified by the ATA you either have to produce a diploma or something or, to prove a minimum of 5 years experience, INVOICES for jobs accomplished. Why can't ProZ.com be a little more like the ATA? All problems solved.
Invoices are not required for ATA certification. See here:
http://www.atanet.org/certification/eligibility_requirementsform.php
All you need is to prove that you have 5 years' experience, and you can prove it by providing letters of reference from past clients (and if you have academic qualifications, you don't even need to prove 5 years' experience). Invoices are one of the suggested options for people who are unable to provide letters of reference.
And you can be an ATA member with all the benefits that matter without being certified (as I am).
| | | | | I agree with Nicole | Jul 3, 2012 |
I don't believe any voting system would work. Also, should we distinguish between British English, Irish English, American English, Australian English, etc. Those differences might be very important if the translations are to be published by a publishing house in a particular country, or if they are intended for publication in the most prestigious papers. What about English for international audience, who has problems understanding, especially, idiomatic British English?
All certa... See more I don't believe any voting system would work. Also, should we distinguish between British English, Irish English, American English, Australian English, etc. Those differences might be very important if the translations are to be published by a publishing house in a particular country, or if they are intended for publication in the most prestigious papers. What about English for international audience, who has problems understanding, especially, idiomatic British English?
All certain associations want is money, not ProZ -- they are not an association and the money is just for membership -- this is fair. Many of translator certification programs are just so that the association makes additional money. This is why they agree to take experience as proof of professionalism.
The only honest and reliable exam, which is really hard, is the court interpreters exam some courts require in the US, because nobody really makes any money on that. This is the only reliable exam I know.
[Edited at 2012-07-03 20:39 GMT] ▲ Collapse
| | |
|
|
|
Samuel Murray wrote: Charlie Bavington wrote:
I just want to see false claims removed about the level of service a profile purports to represent. That is fine, Charlie, but those who pass your test should not be allowed to say that they are verified as native speakers, but rather those who fail your test should be allowed (forced?) to say that they are verified as being not native speakers. If the aim of a test is not to find honest translators but to find lying translators, then any certificates awarded or labels granted in terms of the test should be given to the liars, e.g. a little icon on their profile saying "verified non-native".
I made a similar if flippant suggestion a while ago. It was thought draconian. I can see that.
(And yes, absolutely, the emphasis is on failing the test; passing it is not = yellow "N")
Given that this issue is related to misrepresentation (lying!) for the purposes of securing paid projects, perhaps it could be highlighted in some other way, while leaving the current sytem of yellow "N" and black "N" unchanged (your own efforts towards improving yellow "N" notwithstanding), and except for not allowing profile owners to make a fresh claim for an "N" of either sort for a number of years, relating to the "working pairs" list.
For example, a different colour for "working" pairs where the target language is known to be definitely non-native (and while we're at it, perhaps also a colour for pairs where the "N" is yellow, the rest being simply an unknown quantity, or indeed quality). Perhaps green (verified), amber (unknown) and red (known to definitely not be a native language although it was previously claimed as such) would be clear? I this way, I see your (and others') focus to have been on the greens thus far; mine (and others') on the reds.
Those with a red pair would also be blocked from applying for projects in that pair.
[Edited at 2012-07-03 20:55 GMT]
| | | | | Thank you Charlie | Jul 3, 2012 |
Charlie Bavington wrote: Samuel Murray wrote: Charlie Bavington wrote: Michele Fauble wrote:
I would propose that this be limited to linguistic characteristics, since it is these that are relevant to translation quality. And this threshold would indeed relate purely to the quality of written output. Well, I don't want to "disagree", but I do want to "voice an opposite opinion": I'm sure nativeness or non-nativeness can be tested using a written test, ..... I realise that speech is not nearly as important to translators as written aspects of the language, but we're not testing how good you are in a language -- we're testing "how native" you are in that language, I need to pop out, but briefly - testing translators speech is IMHO pointless, and time consuming. The whole point of translation is that it is written. One could easily be deaf (at least one active Fr-Eng kudoz contributor was, back in the day), or mute. Extreme perhaps. But possible, and demonstrativ of its relevance, i.e. borderline to none. I think the difference is this. You appear to want to comprehensively test a particular attribute personal and specific to the individual that owns a profile. You want positive proof they are who and what they say they are. They need to earn the "N". I just want to see false claims removed about the level of service a profile purports to represent.
Thanks a lot for thinking to deaf people !
I really think that this thread is sinking deeper and deeper into the murky waters of segregation, denunciation, spying, intrusion in privacy (far from your initial purely professional objective), etc.
I even read somewhere in this thread that some colleagues could act as "judges" (sic)!!
It's getting nauseating.
Don't you think that if the huge amount of energy which has been expended in this thread since the begining had been used to find new clients, it would have been far more useful?
Catherine
| | | | Samuel Murray
Netherlands
Local time: 05:03
Member (2006)
English to Afrikaans
+ ...
| Next: how do find candidates for the test? | Jul 3, 2012 |
Charlie Bavington wrote: Samuel Murray wrote:
If the aim of a test is not to find honest translators but to find lying translators, then any certificates awarded or labels granted in terms of the test should be given to the liars, e.g. a little icon on their profile saying "verified non-native". I made a similar if flippant suggestion a while ago. It was thought draconian. ... And yes, absolutely, the emphasis is on failing the test; passing it is not = yellow "N".
So the next question is how to win candidates for the test. I can see two options here:
1. Make it mandatory. In other words, ProZ.com designs a test with e.g. 1000 questions and lets everyone write a test with e.g. 100 questions from those 1000 questions. The test can be done online and can be marked by a computer. Those who refuse the test are demoted to grey icons (if they don't have grey ones already). Those who pass the test get to keep their current colour icons (whatever it is). Those who fail the test get a red icon. Anyone can appeal a decision.
2. Use whistle-blowing. If enough complaints come in for your nativeness (say, 10 complaints), you get an e-mail saying that you must take the exam. If you refuse, you get demoted to gray. If you pass, you keep your colour. If you fail, you get red. All whistle-blowers must be prepared to write the exam as well (and if they refuse, their complaint is simply not counted).
This would only be really useful if the "Native language" option in the search function would be changed to "Any | Language A, verified | Language A, not verified | Language B, verified | Language B, not verified", so that there is an incentive to get verified.
I would probably refuse to take the test, until it becomes profitable not to. For those readers who didn't follow this subthread, please note that I don't favour the idea of a written test, but there is nothing wrong with discussing how a written text would have to work if implemented.
[Edited at 2012-07-03 21:30 GMT]
| | | | Nicole Schnell
United States
Local time: 20:03
English to German
+ ...
In memoriam | Hooray, the witch-hunt! | Jul 3, 2012 |
Samuel Murray wrote:
a test with e.g. 1000 questions
2. Use whistle-blowing.
This will certainly please the candidates who display a pompous, 1 inch wide "Certified Pro" logo on their profile page as if it were some linguistic degree. Neat way to deceive potential customers.
I would probably refuse to take the test, until it becomes profitable not to.
Good thinking.
| | |
|
|
|
| Different approach again, old bean | Jul 3, 2012 |
Samuel Murray wrote:
....so that there is an incentive to get verified.
I'm not coming at this from the angle of incentives to get verified. I just want a disincentive to lie. I thought we agreed earlier today, when we both posted that we wanted to be able to take profiles at face value (mine was headed "innocent until proven guilty"; you used a similar phrase). I don't need people to prove every last thing. Just that they know that if they ARE caught lying, they may as well delete their profile, because it will be worthless.
Hence, even if it didn't sound an expensive nightmare to run, I'd discount your option 1.
Some kind of report function would be the first step, I think (possibly not unlike the Pro one about which I knew nothing before today but sounds similar to what I think could be useful, although I take the point about the climate it might create). Nothing systematic, just as and when people spot situations.
Clearly (to me!) it would need to involve the opinion of several people (but how to select them?) and clearly, there would need to be some kind of way to appeal or justify oneself. And the sort of grounds allowable - posts on here? elsewhere? kudoz questions and answers?
I'm not over keen on the idea of written tests because in this context, cheating would be ridiculously easy (or preventing it prohibitively expensive). I see no reason to exclude supporting documentary evidence at this stage, voluntarily (as I said before, as evidence not proof). I like Sheila's idea of IM/chat - 2 minutes should be ample, failure to reply instantly being viewed suspiciously. Again, you'd need to have a few one-on-ones I think, an odd number, to avoid a tie. Those involved in this would have to not work in the same source-target pair as the, er, "accused" (?!)
Combine all that with the green-amber-red and consequences of failing (lying) I outlined not long ago, and you've pretty much got the sum total of my conclusions thus far
| | | | | I have a brilliant idea | Jul 3, 2012 |
After reading all of those shocking ideas that professional people should be tested, spied on, categorized, put into boxes, prejudiced against, there should be just one option in the profiles - native language - private, or not to be disclosed before bidding. Everyone can inform the client what his or her native language is while bidding on the job.
| | | | Andy Watkinson
Spain
Local time: 05:03
Member
Catalan to English
+ ...
LilianBoland wrote:
Also, should we distinguish between British English, Irish English,
American English, Australian English, etc.
You sound like you're asking, but I'll assume it's a statement.
Well, Lilian, we've all been taking this into account for decades.
You're right; just a bit late and this is NOT the topic.
Those differences might be very important if the translations are to be
published.... or if they are intended for publication in the most prestigious papers.
Again Lilian, we know.
What we don't know is why your perpetual point of reference should be newspapers, especially those in the US.
All certain associations want is money, not ProZ -- they are not an association and the money is just for membership -- this is fair.
Yeah, right.
| | | | Nicole Schnell
United States
Local time: 20:03
English to German
+ ...
In memoriam | No matter what this all is about | Jul 3, 2012 |
LilianBoland wrote:
After reading all of those shocking ideas that professional people should be tested, spied on, categorized, put into boxes, prejudiced against, there should be just one option in the profiles - native language - private, or not to be disclosed before bidding. Everyone can inform the client what his or her native language is while bidding on the job.
Contributors from other countries than the US, comfortably nested in cutesy and comfy EU laws and regulations apparently have no clue what they are talking about in terms of background checks and privacy laws.
| | |
|
|
|
Andy Watkinson
Spain
Local time: 05:03
Member
Catalan to English
+ ...
LilianBoland wrote:
After reading all of those shocking ideas that professional people should be tested, spied on, categorized, put into boxes, prejudiced against, there should be just one option in the profiles - native language - private, or not to be disclosed before bidding. Everyone can inform the client what his or her native language is while bidding on the job.
"...spied on, categorized, put into boxes, prejudiced against, ..."
Please, this description has nothing to do with the proposals put foward here.
This process is called peer review and is firmly established as standard practice in an enormous range of professional/academic/research etc fields.
It also happens to be a less labour intensive option for the site (making it a more probable solution, as things stand) and if you're confident that your claim is accurate, well....
| | | | Phil Hand
China
Local time: 11:03
Chinese to English
| At what point does sincere disagreement turn into trolling? | Jul 4, 2012 |
@Charlie
Your "verified non-native" idea has two big problems, as I see it. (1) There's no way the staff will agree, because it's too draconian. (2) It's a threat, and for threats to work they have to be widely understood and feared. Neither would be the case on Proz - new users wouldn't understand it, and lots of people sign up just to see how it goes, without any level of investment in the site. For them, getting marked "pants on fire" or getting chucked off wouldn't be much of a ... See more @Charlie
Your "verified non-native" idea has two big problems, as I see it. (1) There's no way the staff will agree, because it's too draconian. (2) It's a threat, and for threats to work they have to be widely understood and feared. Neither would be the case on Proz - new users wouldn't understand it, and lots of people sign up just to see how it goes, without any level of investment in the site. For them, getting marked "pants on fire" or getting chucked off wouldn't be much of a punishment.
@Robert
You're right that a written test wouldn't be a true test of native competence, but it would be a functional diagnostic test (and one which is more likely to give false positives than false negatives), and that's all we need.
My version of Sheila's idea from a few pages back:
"I think there would have to be a formal, clear mechanism by which second native languages could be accepted. Like you say, it's a bit hard to control, so I'd suggest something similar to a translation test. Proz sends a list of three questions to an applicant generated at random from a massive list that users supply. Applicant has to write 100 words in response to each and reply within 20 minutes."
These responses could be sent out for formal proofreading, and a certain number of errors per 100 words would be a fail. You could only apply for verification in one language once per year. Similar to Sheila's idea, but logistically slightly easier. ▲ Collapse
| | | | | Post removed: This post was hidden by a moderator or staff member because it was not in line with site rule | | 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? | 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 » |
| | Anycount & Translation Office 3000 | Translation Office 3000
Translation Office 3000 is an advanced accounting tool for freelance translators and small agencies. TO3000 easily and seamlessly integrates with the business life of professional freelance translators.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
