This article first appeared in the Microsoft Language Portal. It briefly summarizes the main terminology management methods in the localization process and highlights the differences between the semasiological and onomasiological approaches.
Terminology management in the localization process has changed considerably over the years, both in the methods used and in the different phases of the product life cycle when terminology management takes place. When localization was still in its infancy, basic lists of terms were compiled at a relatively advanced stage of the localization process, often by individual translators and without any involvement from development teams. Over the years such lists evolved into glossaries with definitions, usually created at the beginning of the localization process; glossaries were generally product-specific and not always shared across different product teams. In the terminology management model currently used by Microsoft, most new terminology is identified before localization starts; English terminologists work together with developers and use a multilingual, concept-oriented, terminology database to make Microsoft terminology available to all localization teams and other users at the same time.
Traditional glossaries and dictionaries follow a semasiological (word-to-meaning) approach. Each glossary entry represents a lexical item in a given language and lists the different meanings associated to it. The English entry frame, for example, can denote “a box-shaped space that contains an image”, “a section on a Web page”, “one of many sequential images that make up video”, “a package of information transmitted as a single unit”, etc. In a term-oriented system, synonyms (e.g. blog and weblog) and any variations (e.g. web log) are recorded as separate, individual entries. Glossaries as used in a localization framework are mainly bilingual and product-specific; each entry contains a source term and a definition, usually in English, with an associated target term; translations are repeated for all variations of the English term. If no centralized terminology sharing and verification system is in place, inconsistencies across product glossaries might occur.
Microsoft terminology management, on the other hand, follows an onomasiological (concept-to-term) approach. Each terminology database entry represents a concept and lists the different lexical realizations that concept can take in the source language, usually English, and in one or more target languages. The database entry defined as “the exchange of text messages and computer files over a communications network, such as a local area network or the Internet”, for example, can be represented by the English terms electronic mail, email, e-mail, mail; in Italian, only one individual term, posta elettronica, is associated to the concept and is therefore associated, or “mapped”, to all English terms. A concept-oriented system makes terminology immediately available to all product teams, reduces potential terminology inconsistencies and can help achieve higher standardization in the target language and, as a result, a better end user experience.
Last week I was in Wiesbaden (Germany) for the tcworld conference 2008, the world's largest event in the field of technical communication, to present Working with terminology at Microsoft: a case study. [...]