Discovering conventions in the occluded academic genre of conference discussion session in support of computer science novice academics

In the past three decades, applied linguists in the field of English for Academic Purposes (EAP) have provided indispensable support in preparing novice academics from a vast range of disciplines in their scholarly writing. However, little attention has been given to scholarly spoken genres, particularly conference discussion session, despite its challenging nature and vital role in academic communication. What questions might be asked and how to respond to and persuade the audience are unaddressed questions that often leave novice academics disconcerted. This paper presents the preliminary findings of an investigation that aims to address these two questions for the computer science community. Ten computer science conference discussion sessions have been collected and transcribed so far with half of them initiated by novice academics and half initiated by experienced academics. To validate the comparison between novice academics and experienced academics, measures have been taken to control variables such as conference type, year of the conference, first language of the speaker and length of the session. The data analysis draws on Systemic Functional Linguistics (SFL); questions from audience are categorised in terms of their ideational function (i.e., what research experience or ideas do audience seek from the presenters); presenter responses are analysed in terms of their interpersonal function (i.e., how do the presenters position themselves) using Appraisal Theory (Martin & White, 2005). The preliminary results were then brought along in an interview with a computer science professor as an ‘insider’ for further interpretations. This pilot study found that nine types of questions were posed by the audience and that experienced academics positioned themselves less as a defender of their research products but more as an ‘honest’ researcher in comparison to novice academics. The findings will be tested on a larger dataset in the next phase of this investigation.