ISO Standards for Localization and Translation
By: Sue Ellen Wright (Kent State University) and Dr. Jiri Stejskal (CETRA Language Solutions) - Kent State University - Institute for Applied Linguistics
18 December 2012
Localization today involves ever-widening circles of contributors and stakeholders – from the originators of source materials to project managers, language engineers, translators, reviewers, editors, layout specialists and more, working in two to forty languages and beyond. As the language industry has grown, experts have found themselves confronted with the same kinds of problems encountered by early railroad builders or fire departments. Even if it looks a lot like we are all doing the same thing with a given procedure (more or less), if rail gauge or hose connections are incompatible, collaboration between potential partners proves difficult or impossible.
Industry-based efforts like those of the now defunct Localization Industry Standards Association (LISA) paved the way by introducing the Translation Memory eXchange (TMX) standard, which lives on after its parent’s demise as one of the most widely applied standards in the language enterprise. OASIS standards such as XLIFF (XML Localisation Interchange File Format) also exemplify the validity of standards that have evolved bottom-up out of hands-on industrial environments. Nevertheless, as we have seen with standards like Unicode and SGML, there are advantages to migrating viable standards into the ISO (International Organization for Standardization) environment as standards mature and gain widespread acceptance.
ISO Technical Committee 37 for Terminology and Other Language and Content Resources writes standards for all manner of language-related resources, including terminology management, corpus linguistics, translation and interpreting, along with basic language-related computing issues, such as language codes. In the US, TC 37 activities are housed in the so-called “US Mirror Group,” which functions under the ASTM Committee F43 on Language Services and Products.
The ISO committee’s scope or mission statement calls for the standardization of principles, methods and applications relating to terminology and other language and content resources in the contexts of multilingual communication and cultural diversity. As a consequence, ISO TC 37 is the logical home for standards in the language industry.
Stakeholders in TC 37 standards include standardizers who must define terminology in their work, professional terminologists operating term banks and large corporate resources, and academics creating text corpora. But less obviously, they also serve the needs of Semantic Web developers and most recently, of translation and localization services providers. Most importantly, but perhaps least obviously, the language industry uses some TC 37 standards all the time without even realizing it – TC 37 “owns” the language codes (ISO 639), for instance.
Most recently TC 37 has formed a new sub-committee, SC 5 for Translation and Interpreting. Current documents include:
ISO TS 11669:2011, Translation projects – General guidance was published in May 2012. The document is a technical specification designed to provide guidance for developing specifications for translation and localization projects. Based on the premise that no two projects will be identical, it guides TSPs and LSPs through the process of defining the specifications against which the quality of a language service or product will be judged. Not yet a full standard, the document is subject to review and comment, with the intention that it will evolve into a true standard after a period of experimentation and revision.
ISO 17100, Requirements for Translation Services is essentially a process standard that is currently passing from a Committee Draft to a Draft International Standard, with the potential to serve as a replacement for CEN’s EN 15038 as basis for the certification of translation service providers (TSPs). A recent ballot resulted in many comments, which has given rise to a resolution phase. The procedures involved reflect the great effort needed to reconcile business interests and cultural differences that may exist between different national and regional groups within ISO.
Clients coming from a background outside translation and localization frequently ask about metrics and measures, reflecting a desire for quantifiable data that can be used to demonstrate translation quality. ISO/WD (Working Draft) 14080, Assessment of translations, began as an effort to devise such a metric for assessing translations during the production process. Unfortunately, this project was terminated in May 2012 due to unresolved issues and administrative hurdles, and the question of translation quality assessment is currently on hold. While this remains a lynchpin issue, it is also one of the most problematic. Many in industry are clamoring for simple, transparent solutions, but unfortunately, it is not a simple problem.
ISO 12616:2002, Translation-oriented terminography (terminology work) is an existing standard that is technically outdated. Plans are in the works for the development of a new version starting in summer of 2013.
In terms of concrete application-oriented standards, however, there is probably no other standard that sees day-to-day application in the localization industry than TMX. Along with other LISA/OSCAR standards, this document continues under the aegis of the ETSI Interoperability Standards group, in collaboration with TC 37. The related SRX – Segmentation of Translation Memories, designed to solve some of the issues inherent in TMX, is already under development in ISO.
Another old LISA standard, ISO’s 30042, Termbase Exchange (TBX) has entered a serious revision phase that will no doubt take a number of years to complete. The work group is exploring TBX implementations and recognized inadequacies, with an eye to introducing modalities designed to address working problems without interfering with backward compatibility with many in-place applications. There is also an effort to take a new look at data categories used in termbases and to provide greater accountability as well as flexibility in defining and using data categories in conjunction with TC 37’s ISOcat Data Category Registry.
In the standards development environment, the Linport project is aimed at enhancing the Interoperability for a variety of translation and localization software solutions (see Dr. Alan K. Melby’s article on this topic), with an overall goal of increased flexibility and nimble coordination of translation working environments.
On the US scene, ASTM's F43.03 sub-committee on Language Translation plans to revise and update ASTM F 2575-06, Standard Guide for Quality Assurance in Translation in alignment with the new ISO standards, especially 11669 and 17100. A new effort is underway to draft a separate ASTM standard that would serve as a basis for certification of translation companies.
Critical to all these efforts is the close collaboration that exists with regional and national bodies (CEN, ASTM, BSI, DIN, AFNOR, CSC, SAC, and many others), as well as with industry standards organizations, particularly in the information technology arena (OASIS, W3C, IETF, IANA, ETSI).
|Sue Ellen Wright is a Professor of German and a member of the Kent State University Institute for Applied Linguistics, where she teaches computer applications for translators and German to English technical translation. She chairs the U.S. mirror committee (Technical Advisory Group) for ISO Technical Committee 37, Terminology and language and content resources and has worked actively on such standards as ISO 30042 (TBX). She is chair of the TC 37 Data Category Registry.
|Dr. Jiri Stejskal, founder and president of CETRA Language Solutions, has more than 20 years of experience as a translator, and has also taught undergraduate and graduate language courses at the University of Pennsylvania since 1990. Jiri is a past President of the American Translators Association and he currently serves as Vice President of the International Federation of Translators and Treasurer of the American Foundation for Translation and Interpretation.|