An Ethnographic Study of the Use of Translation Tools in a Translation Agency: Implications for Translation Tool Design

There is often a significant disjunction between how software designers envision the use of their application programs and how the software is actually understood, perceived, and employed by end users. When this disjunction occurs, software tools and technology fail to meet user needs or expectations and do not fulfill business objectives. This study investigates this disjunction in a translation agency. Some researchers have argued that many software system designs fail because insufficient attention is paid during the design phase to the social and cultural context of how and why people actually work with software applications in their organizational settings. These researchers argue that the development of usable and useful software thus depends on receiving organizationally meaningful and actionable information from end users both during the design process and during the refinement and upgrade cycles of the software. Regrettably, this information, which is essentially ethnographic in nature, is not easy to obtain using the normal methods of requirement specification. This dissertation argues that ethnography can make a significant contribution to the success of the design process and serve as a valuable requirement specification method. The research focuses on the use of translation tools by translation professionals at the agency and uses detailed ethnographic description to document the movement of translation objects (texts, data objects) through the organization and between participants by describing the workflows, processes, and tasks that must be performed with the translation tools. The study uses microethnography to investigate the use of specific features of the translation tools and makes recommendations for translation tool design. There is some evidence in the literature to suggest that many software tool designers find it difficult to translate ethnographic research findings into software design decisions. The dissertation proposes a model for analyzing ethnographic findings for the purpose of translation software design.