Technology forces use of incorrect grammar
Thread poster: Kaspars Melkis

Kaspars Melkis  Identity Verified
United Kingdom
Local time: 08:24
Member (2005)
English to Latvian
+ ...
May 17, 2009

I know that many translators are familiar with Eugene Nida's works on translation theory and practice. Particularly, he described how the first Bible translators insisted on a literal word-by-word translation, copying the grammatical structures, and demanded to introduce the passive in the target language that didn't have the passive. It has indeed happened in Latvian where 10 commandments of the Bible were translated in a novel grammatical form (Tev nebūs zagt). Even though this grammatical form is sometimes parodied or used in fiction to create an effect of an absolute injunction, it is never considered a standard Latvian grammar. Such approach is surely too primitive by current translation standards.

But it seems that technology limitations or lack of time and money or rather lack of interest or laziness is again forcing us to use non-standard grammatical construction. Among many other examples, when I call Latvian cellphone info number to learn how many free minutes I have left, the automatic voice system gives an answer – jums ir četrdesmit divi neizmantotas minūtes that is grammatically incorrect because the word divi is not in gender agreement with the minūtes. The correct form is divas.

My own experience in localization projects is that too little attention is given to the target language needs. System developers love quick phrase concatenation and often ignore comments from translators that it doesn't work in a target language and certain adjustments have to be made. I can understand when it is due to budget or time restrictions, but it irks me when project managers display ignorance and disregard warnings from translators and insist that the language has to be accommodated instead. For example, I have been asked to provide translation for the phrase "by you" to be concatenated to "Editing objects" and now I need to convince the project manager that in Latvian such separate passive construction is impossible and both phrases have to be translated as a whole. This approach of translating only sentence parts is strikingly similar to the literal Bible translation practice that was rejected by Nida.

Do you have similar experience in localization? Does anybody have any suggestions how to deal with such situations?

[Edited at 2009-05-17 12:13 GMT]


 

Neil Coffey  Identity Verified
United Kingdom
Local time: 08:24
French to English
+ ...
Two-way adaptation process May 17, 2009

Speaking as both a linguist and a computer programmer, I think the answer to this probably has to be a two-way process, where (a) programmers are educated a little more in designing their message bundles to cope with more cross-linguistic variation, and (b) we as humans learn to be less shocked by this type of "computer grammar".

So in (a), I would mean, for example, not just having the option of "singular or plural" messages, but taking into account some other common paradigms: e.g. allowing for the fact that the form with zero may or not be the plural form, or trying to avoid messages ending in a preposition which the programmer naively appends to other messages.

Perhaps library designers need to extend their notion of "locale" and "language" to be more a more linguistic definition. For example, from the Java programming language (and this situation is typical), you can query the current locale's preferred date format, currency format, currency symbol... but really nothing linguistic. There's no call you can make to ask "should the computer use an honorific form of address?", "should I use the plural form with 'zero'?", "does the current locale's language have morphological cases?" etc. And if there were, would it be worth programmers worrying about localised messages down to this level of detail?

So rightly or wrongly, in (b), I think we have to accept that for practical reasons "localised message grammar" is likely to remain a little different from "human grammar". So when your answer phone tells you "Usted tiene uno mensajes nuevos", one way of looking at this is that it's just a different type of grammar. In human speech, it is ungrammatical, but in the specific world of localised computer-generated prompts, maybe we argue that it's in some way "grammatical".

[Edited at 2009-05-17 22:10 GMT]

[Edited at 2009-05-17 22:16 GMT]


 

Clarisa Moraña  Identity Verified
Argentina
Local time: 04:24
Member (2002)
English to Spanish
+ ...
Tecnology or project managers? May 18, 2009

Or even bad translators selected to review our words?
My experience has told me that when my translation job into Spanish is reviewed by bad project managers who insist to find out the same number of words, and even the same words of the original English, they pressure me to deliver a bad quality job. They do not understand regarding pronouns, nor that in my language the verb informs the subject, and they do not understand many things regarding translation (I insist that the important thing is to convey the same meaning, not the same words, with the right grammar.
While I do understand that every translation should be reviewed, it is important to understand that the final product will be read by the target audience and that we have to take into account that target, who might find, for instance, that the user manual is "ugly" and uses weird expressions (exactly as it was -very well - written in the original.
In my opinion, it is not the technology, but those who expect that the translation be a foreign language mirror of the original.

Kind regards

Clarisa


 

Riccardo Schiaffino  Identity Verified
United States
Local time: 01:24
Member (2003)
English to Italian
+ ...
There are many translation techniques that help in localization May 18, 2009

Neil Coffey wrote:

So rightly or wrongly, in (b), I think we have to accept that for practical reasons "localised message grammar" is likely to remain a little different from "human grammar". So when your answer phone tells you "Usted tiene uno mensajes nuevos", one way of looking at this is that it's just a different type of grammar. In human speech, it is ungrammatical, but in the specific world of localised computer-generated prompts, maybe we argue that it's in some way "grammatical".

[Edited at 2009-05-17 22:10 GMT]

[Edited at 2009-05-17 22:16 GMT]


Although it is more difficult in other instances, you have given a perfect example of a concatenated message where the translator can actually fix most problems:

"You have %d new messages", if translated as "Usted tiene %d mensajes nuevos" creates an ungrammatical sentence in Spanish (or in the original English, for that matter), when the value of the variable is 1.

However, "You have %d new messages" can be translated as "Número de mensajes nuevos: %d", which yields a grammatically correct sentence, no matter how many new messages there are.

Similar techniques can be used in many instances, for many languages, although of course they are not a panacea.


 

Adam Łobatiuk  Identity Verified
Poland
Local time: 09:24
Member (2009)
English to Polish
+ ...
Some good points here May 18, 2009

Riccardo Schiaffino wrote:

However, "You have %d new messages" can be translated as "Número de mensajes nuevos: %d", which yields a grammatically correct sentence, no matter how many new messages there are.

Similar techniques can be used in many instances, for many languages, although of course they are not a panacea.


That's what we do in Polish. It also helps to use abbreviations of units instead of the full words, because there is a different plural form for 2-4 nouns and 5 or more. The above example would work for Polish; however, you'd have to ensure that "two" is a dedicated variable for that string, because it has a masculine and feminine form.

Clarisa wrote:

While I do understand that every translation should be reviewed, it is important to understand that the final product will be read by the target audience and that we have to take into account that target, who might find, for instance, that the user manual is "ugly" and uses weird expressions (exactly as it was -very well - written in the original.
In my opinion, it is not the technology, but those who expect that the translation be a foreign language mirror of the original.


My thoughts and experience exactly. What's more, the English I get to translate, even for high-profile end clients, is often quite sloppy. Plus it seems that English is more tolerant towards repeated words (maybe because it's an analytical language, where many auxiliary words are used?), while repetitions look and sound horrible in Polish. No one would normally keep repeating e.g. "e-mail message" in its entirety sentence after sentence, when "message" (or even "e-mail") would do. Yet, in various"quality assurance reports" such omissions and even inflected forms count as "inconsistencies".


 

Rod Walters  Identity Verified
Japan
Local time: 16:24
Japanese to English
Not technology - economics and ignorance? May 18, 2009

Rather than technology forcing the unfortunate grammar, it's more a case of economics and ignorance.

The people who are responsible for processes which involve language often aren't particularly knowledgeable in that area, and indeed, they often fail to give it a moment's thought in design. So that's ignorance.

When it comes time to fix this situation, economics steps in and forbids it. Fixing the situation (with the thoughtful suggestions above) costs money, and when you buy say, a printer, you want it to be as cheap as possible. Most purchasers seem happy to pay for 'just good enough and no more'.

Certainly in Japan, I've translated manuals and marketing for extremely costly, sensitive technical equipment (boiler systems, scientific microscopes etc.) where the entire interface with the buyer and user has been almost totally overlooked. How much value to everybody that is lost because of this is hard to quantify exactly, but I suspect it's very high indeed.

Perhaps translators need to involve themselves more in process design and help to save everybody money and their beloved linguistic purity.


 

DZiW
Ukraine
English to Russian
+ ...
simplification tRend May 18, 2009

English, including the King's English, belongs to Germanic and many researches consider it to be 'just a simplified version of German'. But most Anglophones cannot anglicize it in all conscience because 'they neither understand, nor can speak German'. So?
I think the simpler the rules and words, the easier to communicate and understand each other. What's wrong about it? Oh, it doesn't look familiar or 'humanish' ?
Almost any intelligent programmer can tell you that machine can cope with the algorithms but not so many irrational, absurd, unreasonable, and illogical (=human) exceptions. Things just happened and we took them in our stride for granted - The History. Why, that is no excuse.
IMO I can tolerate it until it does cause real damage.

CheArsicon_wink.gif


 

Heinrich Pesch  Identity Verified
Finland
Local time: 10:24
Member (2003)
Finnish to German
+ ...
Yes, known problem May 18, 2009

How many times I had to answer customers asking: What is "bis" or "to" in Finnish? Sometimes it's unavoidable to create output that is sub-standard linguistically. How to translate "months" into Finnish if you do not even know where the string will be in the English sentence? It could be kuukautta, kuukausia or kuukaudet depending on the syntax.
Software would be at least as intelligent as a modern machine translation package to be able to parse sentences in foreign languages. So we have to compromise. There are too many rules to cover even the main languages.

Regards
Heinrich


 

Kaspars Melkis  Identity Verified
United Kingdom
Local time: 08:24
Member (2005)
English to Latvian
+ ...
TOPIC STARTER
Thanks and more... May 18, 2009

Thanks to all for responses and good suggestions.

So rightly or wrongly, in (b), I think we have to accept that for practical reasons "localised message grammar" is likely to remain a little different from "human grammar". So when your answer phone tells you "Usted tiene uno mensajes nuevos", one way of looking at this is that it's just a different type of grammar. In human speech, it is ungrammatical, but in the specific world of localised computer-generated prompts, maybe we argue that it's in some way "grammatical".


I have only some experience with PHP/Perl programming but in most cases it shouldn't be too difficult to actually implement different prompts for singular/plural, should it? Latvian also uses singular nouns with numbers ending with 1, except 11. I know some projects where it was actually implemented. Interestingly that the locative case of hours in the time reference always take plural form in all cases – at 1 hours, 11 hours or 21 hours.

Riccardo Schiaffino wrote:
"You have %d new messages", if translated as "Usted tiene %d mensajes nuevos" creates an ungrammatical sentence in Spanish (or in the original English, for that matter), when the value of the variable is 1.

However, "You have %d new messages" can be translated as "Número de mensajes nuevos: %d", which yields a grammatically correct sentence, no matter how many new messages there are.


It really works in many cases. It is easier to deal with complete sentences including all placeholders. My problem starts when developers are instructed to make software to use less memory or for some other reason split the phrase as “You have” and “new messages” to be reused in different places or even between different modules. It is almost always guaranteed to create problems.

How many times I had to answer customers asking: What is "bis" or "to" in Finnish? Sometimes it's unavoidable to create output that is sub-standard linguistically. How to translate "months" into Finnish if you do not even know where the string will be in the English sentence? It could be kuukautta, kuukausia or kuukaudet depending on the syntax.


I can also relate to thisicon_smile.gif I also often get asked why some declined term is repeatedly inconsistenticon_wink.gif

Perhaps translators need to involve themselves more in process design and help to save everybody money and their beloved linguistic purity.


Definitely. Isn't Japanese the language that actually use quite a lot of terms borrowed from English and at the same time very unique counting system that makes the difference between different types of objects (slender, flat etc.)? Must be really complicated to create algorithms to deal with such situations.icon_smile.gif

Another aspect is UI length limitations. Similarly to Russian in Latvian different verbs are used for perfective and imperfective actions. The shorter one will usually be used for UI irrespectively of grammatical considerations. For example, the button "Delete" (a file, etc.) is meant to be a perfective action – pressing it will complete the action of elimination of the file, sometimes irreversibly. The best Latvian equivalent is Izdzēst but most often it appears on UI as Dzēst that can be understood as start of the process of deleting but maybe not completing it yet. Users who have been exposed to irregular UI usage sometimes start to use it outside IT.


 


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


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

Technology forces use of incorrect grammar

Advanced search







Protemos translation business management system
Create your account in minutes, and start working! 3-month trial for agencies, and free for freelancers!

The system lets you keep client/vendor database, with contacts and rates, manage projects and assign jobs to vendors, issue invoices, track payments, store and manage project files, generate business reports on turnover profit per client/manager etc.

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 »



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