Deriving Contextual Defining Information for Technical Terms from a Specialized Corpus - The Case of Kiswahili Health Care Terminology

Abstract

This paper tackles real life terminography problem in Tanzania. Currently there are very few dictionaries on the market to cover definitions of Kiswahili terminology. The problem becomes even more apparent when one attempts to obtain Kiswahili glossaries with detailed information on meaning and usage of technical words. Such glossaries are generally appropriate for non-expert users, i.e., people who have no knowledge of the relevant field but for one reason or another find themselves in a situation where they are confronted with words that are used by experts in the field.

             The effects of the lack of suitable dictionaries and glossaries are that people have no sources to consult in the search for meaning of a technical word that they come across in their daily life. For example, when one goes to a pharmacy to buy medicine, may not be able to understand the Kiswahili descriptions of usage and content that is written on the box. To date, the only available dictionary in the domain of health care is a bilingual English Swahili Dictionary of Medicine, ‘Kamusi ya Tiba (TUKI 2001), which contains only Kiswahili equivalents of English health care terms without definitions.

             Lack of user-friendly dictionaries or detailed glossaries with term definitions and usage information may be due to many reasons, financial and technical, such as but a more specific reason is the problems involved in reasons  obtaining defining information and usage examples that can be useful to a non-expert. This paper is devoted to exploration of corpus data in obtaining information that can be used by lexicographers in generating meaningful definitions and explanations of Kiswahili Health Care terms that are relevant for non-expert users.  It forms part of an ongoing research project into developing domain specific publications that can be used by non-experts in the understanding of technical terms as applied in everyday swahili language.

This article highlights a methodology of using corpus data enhancing terminography work. Here, the corpus the role of a passive expert; assumes that is to say, although the corpus cannot answer direct questions nor generate definitions, it carries useful information in the field of Health Care. The article should be seen as work-in-progress; and as such, it  is not meant to give the obvious answers; instead we demonstrate a practical approach that can be strengthened and used in developing Kiswahili terminography for non-expert consumption.

One important that can be posed her is what should be the relevant term information for the non-expert users and can a domain-specific corpus data provide such information for the terminographer.  Traditionally, definitions have tended to be short, technical and not adhering to a proper sentence structure.  Such definitions are very useful in certain communicative environment, such as expert to expert communications. However, where a layperson is involved the short, non-complete, definitions are difficult to grasp. According to Pearson (1998) term definitions for non-expert users should be different from the ones meant for expert users. Whereas definitions for experts are supposed to be short and technical, definitions for non-expert are expected to be simple, detailed with fewer technical words and more general language words. Based on experience from previous dictionary projects such as the Cobuild  project, long full-sentence definitions are easier to understand for non-experts. These long definitions are based on traditions of language teaching, and mimics how new terms and words are introduced through paraphrasing in a classroom. Interestingly, the British National Health Care System (NHS) has adopted these full-sentence long definitions in their online terminology dictionary, which shares with the end users of this project, as the definitions are intended for the layperson (http:www.nhsdirect.ihs.uk). Our assumption in this paper is that non-technical definitions can be obtained from specialised texts that have been written for the consumption of non experts.