Discourse and Corpus Analysis in ESP and ELF: A Research Case Study of Difficulties in Classifying Spoken Discourse = Discourse and Corpus Analysis in ESP and ELF: A Research Case Study of Difficulties in Classifying Spoken Discourse

다국어 초록 (Multilingual Abstract)

This paper addresses considerations that need to be made when the researchers were classifying samples of spoken discourse as being indicative of English for Specific Purposes (ESP) or English as a Lingua Franca (ELF), based upon observations and data collection of naturallyoccurring workplace English nursing discourse from four distinct Asian locales and problems endemic to accurate classification. Since a number of nursing students at the researchers` university in Japan wish to pursue English nursing studies elsewhere in Asia, observation of naturally-occurring nursing English was carried out for two full working days at four distinct Asian locales (National University Hospital, Singapore, Vicente Sotto Hospital, Cebu, The Philippines, St. Paul`s University & Hospital, Manila, and Gleneagles Hospital, Penang, Malaysia) in order to note nursing English discourse patterns not common to American or British-Englishes, and thus possibly be identifiable as examples of an emerging Asian ELF. It was initially hoped that the results of this research would aid materials designers in developing English learning materials for Japanese nursing students by identifying and incorporating elements of authentic spoken Asian ELF/ESP forms used in nursing English. However, the researchers soon realized that accurate classification of such speech data as being ELF or ESP-specific can be exceedingly problematic. Based upon observations made after collecting and analyzing the naturally-occurring nursing English data, six primary category/classification problem areas were identified：1. Local variety 2. Medical domain specific 3. Speech event specific 4. Idiosyncratic style 5. Spoken grammar form 6. Error (both native and non -native). Considerations and questions regarding accurate classification were included for each of the six problem areas.