Re: [atlarge-discuss] Translation issues

[Jeff Williams <jwkckid1@ix.netcom.com>]

Judyth and all stakeholders or other interested parties,

espresso@e-scape.net wrote:

> At 21:37 -0700 2002/08/17, Jeff Williams wrote:
> >  As I said above your remarks here, these tools can "do much of
> >the work", but certainly not all.  Hence, as a very useful and
> >broadly used tools translation facilities are very helpful.  Of
> >course proof reading would be needed to correct any small errors
> >or other more verbose ones after such translations have been
> >done...
>
> Verbosity has nothing to do with it, Jeff. The best translation tools available fall into two categories:
>
> 1) computer-assisted translation programs, which work magnificently for technical documents written with a tightly-constrained vocabulary and a very plain style (e.g., the Université de Montréal's Météo program which automatically translates weather-forecast data), failing which human terminologists and editors have to translate the material into the form the program understands;
>
> 2) translation memory programs, which in effect match words or expressions in a document against a database of previous translations; the results of processing a document with a TM depend very much on the extent and quality of the database, and at best require a lot more than "proofreading" for "small errors" -- more like a systematic line-by-line edit of the material which, in my experience, takes almost as much time and effort as translating the whole thing in the first place.

  Yo are right that this is one such category and that the accuracy of the
comparison database is if primary interest.  However your conclusion as
additional personal opinion is not broadly shared.  For instance this
type of translation software system is what is used by the UN, and works
very well for them...

> The exceptions would be things like spec sheets for 47 models of the same type of equipment, formulaic correspondence  containing only grammatically simple sentences, etc.

  Yadda, yadda, yadda...   This is not only not a good comparison it is not
even in the same category either...  Such analogies/comparisions as a
argument in response for something you obviously seem to be adverse to
does not lend itself to being a very strong argument on it's merits or based
on this argument/comparision just above that you have not so kindly provided...

>
>
> A human translator whose accuracy falls below 98% is likely to cause the client major problems. A TM-assisted translation which is only 80% accurate is quite likely to confuse or misinform or irritate the reader over the other 20% -- hardly a good outcome, and no way to persuade speakers of the target language that we really understand them.

  I have several years now of using several different types of software/system
based translation facilities.  They have been a watershed of a useful tool
in reducing the human impact of translation of a host of different documents
both of the legal and intricate variety and of the less specific or intricate as well..

>
>
> >  I wasn't referring to "Babelfish" as and example.  I would not use >"Babelfish" for verbose documents or translations...
>
> I'm beginning to think you're using "verbose" in a sense other than its usual one -- that is, for any document with more complex content than a straightforward business letter.

  Yes the term "Verbose" in the obvious context I am using it is directly equitable
to your "more complex content than a straightforward business letter".  And much
shorter and frankly to the well versed or advanced reader, simpler to understand
and use...

> If so, it's true that even the most lucidly expressed, non-legalese constitution and bylaws would qualify as verbose.

  No, not necessarily.  Some bylaws are quite simple in the structure and use of
language and terms.  ICANN's is one such example, BTW.

> However, I believe any translation of such materials needs to be at least 99% accurate and also read smoothly in the target language; otherwise, one shouldn't be surprised if native speakers of that language lose interest in reading them, rather than find themselves inspired to join us.

  I agree as close to 100% as is possible is what should be the goal and at least
a 95% accuracy would be minimal...  Hence using a translation tool to do most
of that work getting at least 80% without human intervention, is a huge aid
or assistance.

>
>
> >As a TOOL there are many good, broadly used, and more than basically
> >adequate translation TOOLS that can be used to do a large, but not complete
> >amount of the translation of documents to various languages.
>
> I presume the "TOOLS" you mean are things like Trados, DéjaVu, Star Transit, etc.

  No, those are pure garbage and not worth using...

> -- in short, the translation memory programs which are widely promoted and sold to businesses (including translation agencies) as a way of slashing costs. Many of my colleagues do use these things, generally because they deal with corporate clients producing highly repetitive documents, and perhaps seeking to cut translation costs by paying only once for each term used if they do not know a second language well enough to realize that the same expression in English may have to be translated different ways in different contexts.

  Bylaws in english is usually easily understood by all native english language readers
or speakers.  Hence why in almost every case that I have seen in 23 years of experience,
I have not seen more than one version in English, French, Spanish, or Portuguese.
Same is true in may experience in Legal contracts many of which are quite verbose.
For the purposes of ICANNATLARGE.COM I believe that the same experience
will be replicated, and hence any dialects in any of those languages as well
as host of others will be very unlikely save Chinese...

>
>
> From what I've seen produced for these clients with a TM and the proverbial quick once-over to eliminate the obvious howlers, the results are sometimes good enough for the purpose (as in the American businessman's propensity for "good enough documentation" as opposed to "good documentation"), it's no wonder people in other countries consider themselves better educated.

  I am sure many form other counties do consider themselves better educated.  I
know that I have often met them...  Unfortunately as an american and the land
of the most overall advanced technologically as well as business country ever
known to man, I usually take such thoughts or expression there of as more
bluster than substance...  However there are individual acceptions from
time to time...

> I also hear frequent complaints from fellow-translators that since their consciences don't allow them to turn in inferior translations, they end up putting in many extra unpaid hours to turn "good enough" into "good".

  I suppose your concern here is really a argument of what good is as compared
to good enough...

>
>
> >> >  I can get just about any kind of document translated into Chinese >(Several dialects), Japanese, French, German, Spanish, Portuguese, Polish >and of course Russian.
> >>
> >> That is excellent ... as long as you mean they'll be translated into something native speakers would recognize as good usage in their own languages.
> >
> >  Yes of course I do mean that such translations would be plainly and >easily recognizable to native speakers of those languages...  Anything else >would simply not be worth mentioning or considering obviously...
>
> Sorry to nitpick but I think there is a big difference between my "something native speakers would recognize as good usage" and your "plainly and easily recognizable". I am, first and foremost, a French-to-English translator; English is my native language and I've worked hard for many years to hone my skills so as to produce a given type of material with the style and vocabulary appropriate to the destined readership. When pressed by a client in a desperate hurry, I can produce a "recognizable" translation from English to French; however, if the material is intended for anything other than internal use, either the client or I will hire a well-educated native speaker of French to turn the result into *good* French.

  As one that has many relatives in France, and once married to a linguist that spoke
9 languages as well as was a UN translator, the French have a excessive compulsion
for various forms of the French language when not spoken by a real frenchman...
Although I myself can read a french newspaper without too much trouble, I would
in no way call myself bilingual as french as a second language.  Yet I have
noticed on Jefsey's France@large Mailing list, now dormant, that much of the
french used there by frenchmen/women was far less that what you seem to
call "Good"...  And that forum was a public forum.  Hence I can therefore
only reasonable conclude that perhaps you are overreaching a bit here...

>
>
> This is not intended as a "flame". I do understand that most of the people on this list are not translators or editors and are probably not as concerned with things like good usage as I am. That being said, though, if we are serious about establishing a worldwide organization and eager to communicate our message to speakers of other languages, I don't think providing rough approximations will do.

  I don't thinks so either.  NOR was I suggesting or ever stated such.  I am only
suggesting that translation software tools are and aid in reducing the load on this
burgeoning organization with proofreading being reduced and turnaround on
documents as something that is produces a "Good" translation...  Hence either
you misunderstood what I have been saying, or?????

>
>
> For every successful ad campaign aimed at a non-English-speaking market, there are at least a dozen *avoidable* failures. If this organization-to-be is to succeed, I think we need to learn from those failures.

  Many failures of a minor sort are usually avoidable but yet not avoided.  The reasons
for lack of that avoidance are many and varied of course, but not particularly damaging
in the long or even the short run...

> We have amongst us people who speak many different languages at home but write in intelligible English here.

  Hummm?  I thought you stated or have been arguing that JUST "intelligible"
is not good???  Or did I misunderstand you on that???

> I hope they won't be offended if I state the obvious: that few people who have English as a second language write it flawlessly, and that sometimes using a single word wrongly can cause no end of misunderstandings within a group with a common goal, let alone when the group is an aggregate of people from different cultures with different agendas.

  Yes such minor mistakes in the use of language can cause such problems.  But
those instances are rare...

>
>
> I could easily make myself MISunderstood in German or Spanish but if I want to be understood, I'd write in my own language and trust a good English-to-German or English-to-Spanish translator (not a computer program, since no program extant really understands human language) to get my meaning across.

  Your contention that "no program extant really understands human language" is
a personal opinion that is not broadly shared and therefore is of limited value...

>
>
> Opinionated as ever,
>
> Judyth
>
> P.S. Since French, German and Spanish are amongst the languages you cited, I'd be very interested to see the results of pumping this message or Sotiris's report of the ALAC conference call through your translation program. Perhaps Jefsey, Hans and Gabriel would have something to say about the results.
>
> ##########################################################
> Judyth Mermelstein     "cogito ergo lego ergo cogito..."
> Montreal, QC           <espresso@e-scape.net>
> ##########################################################
> "History teaches us that men and nations behave wisely once
> they have exhausted all other alternatives." (Abba Eban)
> ##########################################################
> See the UNESCO OBSERVATORY ON THE INFORMATION SOCIETY!
> http://www.unesco.org/webworld/observatory
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: atlarge-discuss-unsubscribe@lists.fitug.de
> For additional commands, e-mail: atlarge-discuss-help@lists.fitug.de

Regards,
--
Jeffrey A. Williams
Spokesman for INEGroup - (Over 127k members/stakeholders strong!)
CEO/DIR. Internet Network Eng/SR. Java/CORBA Development Eng.
Information Network Eng. Group. INEG. INC.
E-Mail jwkckid1@ix.netcom.com
Contact Number: 214-244-4827 or 972-244-3801
Address: 5 East Kirkwood Blvd. Grapevine Texas 75208



---------------------------------------------------------------------
To unsubscribe, e-mail: atlarge-discuss-unsubscribe@lists.fitug.de
For additional commands, e-mail: atlarge-discuss-help@lists.fitug.de