Obscure Test Translation Evaluation
Thread poster: Lingua 5B

Lingua 5B  Identity Verified
Bosnia and Herzegovina
Local time: 09:01
German to Serbian
+ ...
Oct 5, 2009

Hi all

A fellow translator did a translation test and got his results back saying he failed. He also got an error report file which he asked me to look into....

First of all, IMO, all of the errors were stylistic, and he failed because of the style and tone ( no, it's not either poetry or literature)

Secondly, some of the errors indicated accuracy issues, but did not affect the accuracy of the message whatsoever.

Thirdly, this is the best part.. some of the errors were indicated as "wrong preposition", while there was no any error at all related to any preposition. The errors labeled "wrong preposition" were either a "wrong" phrasing, wrong term ( though I wouldn't call it wrong, just synonymous and completely fine in the context), but the crucial point here is that no preposition at all was affected, which screamed " this editor/ evaluator hasn't got a clue what a preposition is "

My question is : where do they pick evaluators unfamiliar with basic parts of speech, such as prepositions, that are taught in the elementary curriculum of the 6th grade? That'd be like having an orthopedic surgeon who doesn't know what the scapula or the sternum is and despite of that still trust his/her work and evaluation.

What is the true credibility of such tests having linguistically unaware/uneducated editors on the other side?

Any insights are welcome.

[Edited at 2009-10-05 12:21 GMT]


Kevin Fulton  Identity Verified
United States
Local time: 03:01
German to English
Benchmarks incorrectly applied Oct 5, 2009

From time to time agencies ask me to evaluate translation tests according to the SAE J2450 metric. I also apply a number of personal criteria: terminology that must be used for which there are no acceptable synonyms (spark plug or in the case of a diesel engine, glow plug), and terminology for which a range of synonyms can be acceptable (wire, lead, line, etc.). I allow phrasing that may deviate from a strict translation of the text as long as the sense is the same ("the hammer and the wrench are used..." instead of a more literal "the hammer, together with the wrench ..." )

In some cases, I suspect that high-scoring translation tests are used as benchmarks by non-linguists to evaluate further translations of identical source texts. If the benchmark uses one turn of phrase, an equally valid alternate expression may be marked as an error. This may have been the case with your colleague.


Tim Drayton  Identity Verified
Local time: 10:01
Turkish to English
+ ...
It is translation company's problem Oct 5, 2009

It is the translation company's problem if they cannot appoint competent evaluators. Not all translation companies are competent or professional. If your friend is confident about the quality of his work then he should not be too disheartened and keep looking for work.


Volodymyr Kukharenko
Local time: 10:01
Member (2009)
English to Ukrainian
+ ...
The evaluator is just another linguist, and (s)he may be less qualified than the translator Oct 5, 2009

... or even worse, that could be the client himself who is not a linguist at all. Sometimes it happens, and we cannot avoid that.

Such things happened to me a couple of times. And when I had a chance to look at the corrections, I always responded to those which were not justified.

Once I got the feedback that I should "translate it from the scratch", so I asked for the list of corrections. When I got that list, I was somewhere in the middle between hysterical laugh and uncontrolled anger, because not a single real mistake was found, and the reviewer himself made several mistakes. So I decided to cool down a bit and next day I analyzed each correction, provided a number of official and industry-specific references which proved my point in each case. Then I sent my comments to the client, and and next day I got apologies from them and promise "never to use that reviewer again".


Anne Kjaer Iversen  Identity Verified
Local time: 09:01
German to Danish
+ ...
ask for a 2nd opinion Oct 5, 2009

No question about it.

Ask (read: demand) a 2nd opinion.


Kemal Mustajbegovic  Identity Verified
Local time: 16:01
English to Croatian
+ ...
My humble presumtion... Oct 5, 2009

I'm thinking of Bosnia & Herzegovina with three "official" languages - Bosnian, Croatian, Serbian. The question I'd like to ask is: "What are the languages he was involved in?".

E.G. if any of these languages and English are involved the prepositions are the real problem for unskilled/inexperienced linguist.

[Edited at 2009-10-05 14:07 GMT]


Umang Dholabhai  Identity Verified
Local time: 13:31
English to Gujarati
+ ...
End user has to understand Oct 5, 2009

I totally agree with Vladimir. I too have gone through such a nightmare and have always made it a point to give water tight references to prove my point. It takes time but it increases my credibility before my agency who in turn have a valid point to make before the client. This is so because sometimes the clients (end-users) have their in-house reviewers.

There was an instance when such a reviewer painted the whole document red. I wondered for a moment if I was back in my school. The corrections were mainly stylistic ones (in a technical document) and I informed my agency emphatically to get back with a back translation of the reviewed document which I was sure would have been a totally different cup of tea.

I sometimes suspect that some clients make use of their in-house reviewers to modify the source, and have the newer version without an extra cost. I suspect, I may be wrong !


Lingua 5B  Identity Verified
Bosnia and Herzegovina
Local time: 09:01
German to Serbian
+ ...
Thanks Oct 5, 2009

Thank you all for the replies..

I have just talked with him on msn and he says that they initially asked him to do a 800-word free test translation, but then he refused saying it was too long for a free work, and then got this piece of 350 words which he did and failed.

Having this information now, I suspect it could be one of those agencies that is into fake -testing business ( fraud)? Could the entire test translation revision be a farse ( i.e. not real), because that's certainly what it looks like to me? To me, his test translation looks like a handy little piece which could be sold away for good money.

Is there a point in sending them back-revision of the revision if that's the case?


Laurent KRAULAND (X)  Identity Verified
Local time: 09:01
French to German
+ ...
Why not? Oct 8, 2009

Lingua 5B wrote:

Is there a point in sending them back-revision of the revision if that's the case?

Your friend doesn't have to tell the agency that it was you who did the back-revision (they will suspect him of having searched a "favourable" revisor anyway). It would provide the 2nd opinion Anne thinks your friend should demand. As per translation tests (or test translationsicon_smile.gif), that kind of behaviour coming from an agency spells out ***AVOID THEM AT ANY RATE***...

[Edited at 2009-10-08 07:21 GMT]


To report site rules violations or get help, contact a site moderator:

You can also contact site staff by submitting a support request »

Obscure Test Translation Evaluation

Advanced search

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 »
Wordfast Pro
Translation Memory Software for Any Platform

Exclusive discount for ProZ.com users! Save over 13% when purchasing Wordfast Pro through ProZ.com. Wordfast is the world's #1 provider of platform-independent Translation Memory software. Consistently ranked the most user-friendly and highest value

More info »

  • All of ProZ.com
  • Term search
  • Jobs
  • Forums
  • Multiple search