Knowledge management in the professional services: lessons from functional linguistics

This paper makes the case that the linguistic discipline of pragmatics is useful in developing the next generation of knowledge management applications, especially in professional disciplines such as legal services, health care, and education. The four classes of knowledge-based software to be considered-agents, browsers, indexes (ontologies), and user models-are assessed by applying basic concepts from pragmatics to identify critical issues of context, meaning and location that affect usability and utility. This paper considers issues of common interest to software developers and practicing linguists or ethnographers, such as (a) feature density and preference determination in user modeling; (b) contextualization in high-performance and iterative DB queries; (c) social agency as related to such programming languages as KQML; and (d) disambiguation of meaning with natural language ontologies in large corpora. Three points are asserted and defended: first, current interfaces, databases, and query software have not achieved their full potential for capturing and storing information, and rendering it useful for practical tasks. Second, this failing originates in a reductionist view of language. A richer, context-bound analytical framework is required, such as that offered by pragmatics. Third, combined with the practical methods of ethnography and discourse analysis, a pragmatics-based perspective can support creation of the next generation of computer interfaces and databases, as well as advanced search and navigation tools