ATA Member Feedback on Language Standards

Below is a compilation of comments received following my call on the ATA LinkedIn group to solicit feedback on language standards for a presentation at the LISA Open Standards Summit:

Jiri Stejskal • I will be speaking on behalf of ATA members at the LISA Open Standards Forum in Boston on February 28 – March 1, 2011, and I would like to hear your thoughts on language standards. The Summit will focus on developing and implementing a concrete plan of action to address the problems with the current standards landscape. My contribution to this Summit will be based on your feedback and ideas, so please let me know what you think needs to be done in this area. Thank you!

Susan Welsh • What are language standards? I read what’s on the lisa.org link, but find it hard to fathom what they are talking about. It doesn’t seem to have much to do with “language standards,” but more with software interoperability.

Jiri Stejskal • These could be translation quality process standards such as ASTM F2575 or EN 15038, interpreting standards such as ASTM F2089, translation quality assessment standards such as SAE J2450, or terminology-related standards (there are a number of such standards on ISO level). LISA is typically more involved with technology-oriented standards, but wants to hear from ATA on what standards are important to us and why, what standards should be developed, and get a sense if ATA members care about standards to begin with.

Sinem Canım • I am one of the new members of ATA. I am working in Turkey as a freelance translator and an academic scholar at Istanbul University. I don’t know whether my observations about the case of Turkey in terms standards will be helpful.
First, there is no standard to regulate how much and when to pay to the translator. Sometimes you receive the payment in two weeks sometimes more than one month. Again, there is no standard about rates. Some companies find your rates too high that they do not want to work with you. On the other hand, I hear from my students that they translate 200 words for just 2-3 dollars. When I tell them that this is a very low amount, they tell me that they need money to go on their education. I agree with them in that they cannot be so selective during their course of study. But this shouldn’t result in making use of young labor.

John Shaklee • Dear Jiri: I come at this as a Court Interpreter, but the standards (I believe) would be the same. Translators would comply with a Code of Ethics (a la NAJIT accuracy, confidentiality, limitations of practice, maintenance and improvement of skills and knowledge, accurate representation of credentials, impediments to compliance). Is that what you’re seeking input for? When I do translate (as this past weekend), I follow the same standards. I also believe in a testing body as in the ATA Certification Exam. Just some thoughts. Thanks.

Eileen Hennessy • Definition, please, of the term “language standards”? Each commenter here seems to have a different understanding of what the term means. Since I haven’t a clue what it means in this context, I’ll refrain from commenting.

Dorothee Racette • Dear Sinem,
welcome to ATA! Our members now come from over 90 different countries, so it would be very difficult, if not impossible, to come up with anything but vague recommendations for payment terms. Also, the ATA strictly complies with antitrust laws and does not engage in the discussion of specific rates. The standards Jiri is referring to focus on the difficult question of quality assessment in translation and interpreting.

Eileen Hennessy • It would almost be easier to come up with the recommendations for payment terms. Getting translators to agree on matters of “quality” is equivalent to herding a group of cats. Just to take the case of English, which of the world’s Englishes is going to be considered as the standard in terms of quality? The world’s Spanishes vary even more wildly, to the extent that very often we S-E translators don’t need to look for a place name in order to determine the country from which a text emanates. That’s before we talk about the fact that we translators tends to think our translations are of course better than those done by any of our colleagues. … Major can of worms here!

Elizabeth Hill Barsanti • I don’t know, but it seems to me that trying to regulate how people translate would be like trying to regulate how people speak. How do you set standards for that?

Margarita Griggs • The phrase needs to be more explicitly, not just “Language Standard.” What do you mean? Norm, level, quality, moral principle, pronunciation and spelling of certain language, ordinary or average, etc? I agree with Elizabeth, every translator has his or her own style of writing. Is there “Arts Standards?” Well, translation is an art for me!

Elizabeth Hill Barsanti • and the ethics side of the profession has been thoroughly explored already and standards have been set – see ATA Code of Professional Conduct at http://www.atanet.org/membership/code_of_professional_conduct.php and Proz.com professional guidelines at http://www.proz.com/professional-guidelines/

Martin Cross • There is no point in talking about quality standards outside of the context of fitness for a specific purpose. The standards governing a translation of a document to be used as evidence in court would need to be very different from the standards governing website localization, which would again be different from the standards for literary translation. It is possible to apply standards in terms of production methodology (for example, second translator review, glossary enforcement, etc.) and evaluation (for example, sampling frequency, score weighting, etc.) but each of these would need to be specific to a purpose.

Doris Ganser • A few more considerations in examining potential standards:
Translation for information or publication?
For literary or commercial market?
[ and advertising, which falls in between being mostly pure fiction]

George Rimalower • Jiri, it seems to me that the focus needs to be on processes. I suspect that the best way to measure standards is by evaluating the process that is involved between inception and completion of any specific translation/localization project. Omitting or leaving out vital steps in the process will, in my opinion, compromise the quality of any project.

Ted Wozniak • Whatever standards may be proposed, any approach that treats translation as a “manufacturing process” must be avoided. As we all know, translation is not a commodity and any “production standards” must avoid a “widgets-based” approach.

Personally, I just don’t see a positive cost-benefit ratio from implementing ISO-like standards. All they do is increase costs with little if any improvement in quality (assuming that even minimal QA steps are taken absent such standards).

And under NO circumstances should such standards be promoted as “absolutely necessary” or mandatory. This is still primarily a cottage industry despite the existence of a handful of mega translation companies. Such standards may work for auto manufacturers, microchip producers etc. but we are not in the same type of business.

I’d suggest looking at international auditing standards for an approach that may work for our profession.

Elizabeth Hill Barsanti • What is the purpose of establishing “Language Standards” (?) for translators? To acquire greater professional recognition and credibility for the profession? How, in addition to various evaluation methods such as translation tests, certification, contests, to ‘standardize’ an activity that is, among other things, a craft?

Sue Ellen Wright • I agree with Ted that the imposition of one-size-fits-all standards for translation product and process or even for certification are problematic if all they do is impose additional hassles and costs. By the same token, a major thrust behind the effort to establish a new ASTM committee for language services is aimed at raising the standards for the sourcing of language services by the federal government in the US. There is a legal mandate that all goods and services sourced by the government adhere to existing industry standards. Since no such standards exist per se in the language industry, large government contracts generally go to the lowest bidder without reference to quality output. Leaders in the public sector are thus frustrated by their inability to reward quality and competence. It is this frustration that underscores a move for some sort of standards in the industry. The trick will be to avoid the problems on the one hand while enabling some control on the other.

Eileen Hennessy • My impression is that government contracts of any size, on any level of government, go to the lowest bidder, and that in many instances this is in fact mandated. I doubt that the existence of “industry standards” for the language industry would change the practice.

Martin Cross • NASA contracts don’t go to the lowest bidder.

Standards can make life easier for both producers and purchasers.

I think George is probably right in that standardizing the production processes is the most practical approach. It might also be possible to standardize evaluation processes.

Sue Ellen Wright • With regard to Eileen’s comment: currently language contracts go to the lowest bidder because there are not standards. I am working directly with people involved in procurement of language services in the federal government who want to change that. If you are selling screws (OK, we’re back to commodities, but it’s an easy item to cite), you will not win a contract if it can be demonstrated that your screws to not conform to industry standards. There is a federal statute to this effect. I’ll get back to my people and get the precise clause for you.

Jiri Stejskal • Thank you all for your contributions so far. Let’s keep the discussion going! Lot of good feedback. I will compile and share the results as we get closer to the LISA event at the end of February.

Sue Ellen Wright • Heres some info from Bill Rivers, the chair of the new ASTM F43 main committee on language services:
See http://www.dsp.dla.mil/APP_UIL/content/policy/4120-24.htm and the US National Standards Strategy: http://publicaa.ansi.org/sites/apdl/Documents/Standards Activities/NSSC/USSS-2005- FINAL.pdf (you may need to search for this on the ASNI site – the spaces didn’t copy properly in the URL so not sure this link will work) and the US Cataloging and Standardization Act (http://www.law.cornell.edu/uscode/html/uscode10/usc_sec_10_00002451—-000-.html).

Sue Ellen Wright • More comment from folks I work with who write solicitations for government contracts:
There are an increasing number of people in the Government who are including wording on standards in solicitations. Thus even the lowest bidder would need to implement standards. And contracts do not always go to the lowest bidder.
So the moral is: if you want high standards, you have to write them into your solicitation.

Michael Metzger • I would like to entertain the idea that a language standard should give us a method to measure quality of a translation output with clear definitions of what exactly is measured.
This standard would not be a process quality measurement as they do already exist to my understanding (ASTM, ISO) but something that really focuses on the outcome of the transfer process, the translation product.

Eileen Hennessy • A language standard might be possible for texts that are very limited in content and very cut-and-dried. For any other type…The art of human language is so fluid, shifting, individual, and intangible that, as with all the humanities and liberal arts subjects, the measurement activity that works well in the sciences for tangible products, physical objects, and technical processes can’t be applied to it. Also, whose “standards” of writing style are we going to adopt? Which version of those two major linguae francae of the world, English and Spanish, shall we adopt as the “standard”? What about differences in terminology? in grammar? … Or are we talking about standards for the presentation of the “product”? Margins of x width, so many words on a page, font style Y for legal documents, X for business documents, etc.? That might work–if we can resolve the size difference between 8 1/2 by 11 American paper and A4 European paper, not to mention Chinese paper, Indian paper, etc.–to take one small example of questions to be resolved.

Sue Ellen Wright • Most standards would not touch the kind of detail you are mentioning here, and of course there are different kinds of texts that require different kinds of specifications. Standards are more likely to run in the direction of defining process, requiring certain credentials for translators, and the like. Layout, the choice of variants (British or American English? German spelling system?), etc. are all part of the process for setting parameters that an LSP goes through in negotiating the initial work order with the client. This is one of the problems with setting hard and fast metrics for quality, but overall guidelines for determining parameter-specific metrics could be useful, provided they aren’t too difficult or time-consuming.

Elizabeth Hill Barsanti • … then these ‘Standards’ would be guidelines for assigning translations of tenders and the like at a governmental level (and not involve freelancers, at least not directly). I read tenders of this sort often (in English), from countries worldwide, and they are always translated very correctly. A QC phase is evidently already in place for this type of document.

Sue Ellen Wright • Any standards that might come out of the F43 committee would be generic and not specific to the US government, but part of the motivation for them is being driven by people who source translations and want to be able to require hire standards. We are not talking about translating tenders themselves, but rather assuring quality for language training and translation work that is the subject (product, services) of some of those tenders.

Michael Metzger • Perhaps commenting in a thread soliciting Feedback on Language Standards is the wrong place but I am still more interested to find a method to establish the quality of the translation work (product). If the parameters of the process (ASTM, Quality Assurance) are defined this should be less difficult. So yes a Standard to establish Product Quality would be interesting from my perspective. I believe LSPs do have such internally just as LISA offers the QA Model. What if ATA would come out with a standardized evaluation metric (comparable to an ATA accreditation test) to help us establish the thresholds for unacceptable, good (“enough”) and excellent translation and move from the subjective to an objective common ground of understanding these values?

Doris Ganser • I don’t have time to go into any detail here since I am off on a cruise tomorrow morning with lots of mopping up to do beforehand. Quickly this: I agree with Michael Metzger’s statement that, as LSPs, we do have QA standards but they are liable to be somewhat subjective at times. This applies particularly to the evaluation of translations to more exotic languages, for which we have to rely on outside translators who occasionally compete with those whose translations they evaluate for us. It would be advantageous to have better standards for such evaluations, guidelines that are a bit more precise than our own (meaning mine for my company.) On the other hand, to be more objective, it may be worth mentioning that for years, I have avoided even looking at CVs BEFORE evaluating a translation sample. Yes, Sue-Ellen, normally the translation sample evaluation shows that those with proper translator training provide a higher-value product but it is not always the case. Some outstanding translators have come from other professions (engineering, pharmacists, attorneys), and I remember one who used to be a truck driver (Lee will remember – a later president of ATA.) By the way, Jiri, thank you for starting an interesting discussion.