The Localisation Industry Standards Association (LISA) presents a good global view of localization-relevant standards (LISA 2004), including Cadieux's Automated Localization Workflow Model, which superimposes standards references at critical nodes in his representation of the workflow cycle. Wright and McClure (2003) attempted a comprehensive overview of standards activities, and Wright maintains a retrospective Web site (Wright 2005) with sporadic updates. Of course, the best way to stay abreast with ongoing developments is to follow industry news sources such as MultiLingual Computing e Technology, OASIS's Web-based Cover Pages, and The LISA Newsletter, but maintaining a clear perspective on the advancing positions of the different standards remains a challenge. This problem is exacerbated by the lack of any centralized steering authority or clearinghouse where information about language standards efforts could be posted.

1. Types of language standards

Wright and McClure group language standards into seven major areas:

• Base standards include various markup languages (SGML, XML, HTML, etc.), together with metadata resources, and standards governing character coding, access protocols and data interoperability.

• Content creation, manipulation, and maintenance usually take the form of XML-based authoring standards and standards for text and content markup.

• Translation standards (which in general address localization issues as well as just translation) define parameters for negotiating translation and localization contracts (e.g., DIN 2345, ASTM 15.XXXX), provide metrics for evaluating translations (SAE J2450, ATA Framework, The LISA QA Model), facilitate the interchange and segmentation of translation memory units (LISA TMX and SRX), and specify procedures for certifying translation service providers (CEN EN 15038:2006, Translation services – Service requirements).

• Terminology and lexicography standards, specifically the standards of ISO TC 37 (Terminology and language and content resources), treat a range of topics, from terminology theory to data interchange. Controlled language standards prescribe more restrictive style, grammar, and terminology usage.

• Taxonomy and ontology standards, specifically ISO and ANSI thesaurus standards and the Web Ontology Language (OWL), together with the W3C's Simple Knowledge Organization System (SKOS) standard, are designed to support intelligent information retrieval and ultimately the evolution of the Semantic Web.

• Corpus management standards treat specific markup conventions for logical, syntactic, and semantic markup of text corpora.

• Language and locale-related standards are rapidly progressing to include all the world's languages (and possibly dialects as well), as well as extending to include a range of information specific to the implementation of localization strategies.

2. Standards organizations

Characterization of these standards by type implies a neatly organized field of activity, but in reality a proliferation of organizations and interest groups, both formal standards institutes and various industry consortia, contribute on an ongo-

ing basis to the catalog of language industry standards. Not only is it difficult to maintain a clear view of all these different efforts - the work of different groups also sometimes tends to overlap or conflict, not infrequently resulting in contradictory or competitive standards.