| The construction of stance is of crucial importance in academic language discourse as researchers seek to present their propositions convincingly, build credible argument, and show solidarity with the academic community. This study investigates the construction of stance in reporting clauses in academic spoken registers. The data are drawn from the Michigan Corpus of Academic Spoken English(MICASE), which includes transcripts of academic speech events by native speakers in four main disciplines, namely, Art and Humanities, Biological and Health Sciences, Physical Sciences and Engineering, and Social Sciences and Education. To examine the construction of stance in reporting clauses systematically, a network developed by Charles(2006a) for analysis of reporting clauses is employed in this study. This framework enables stance to be linked with the grammatical and semantic pattern of use of reporting clauses. Quantitative and qualitative analyses of the data leads to identification of differences of construction of stance between academic written and spoken registers and differences among disciplines. Key patterns of reporting clauses in academic spoken discourse are also identified.Results show that there is a considerable degree of visibility of speakers’ role in propositions. In contrast to academic writing where writers incline to conceal their role, speakers tend to take bolder and more direct stance to promote their work and interact with the audience. A key pattern is identified to be frequently used in reporting clauses in academic spoken English: self-source + a personal pronoun + a think verb. Results also suggest that construction of stance differs across disciplines, according to the epistemology and ideology of the discipline concerned. In general, due to the discursive feature of construction of knowledge in Art and Humanities, speakers use more direct and clear-cut stance expressions to make arguments. Whereas in natural science disciplines, namely Biological and Health Sciences and Physical Sciences and Engineering, where researchers make propositions on a experimental and evidential basis, they tend to apply more stance construction strategies to show objectivity. In addition, A few other patterns are distinguished not to be included in the framework of Charles(2006a)(e.g. pronoun you as the subject, it seems/sounds that).The framework of Charles(2006a) contributes greatly to exploring the construction of stance in reporting clauses in academic spoken registers. However, due to the interactive feature of academic speech events and turn shifts of the speakers, it is indicated by results that the construction of stance and engagement are indivisible. Considering the mixed construction of stance and engagement and the patterns identified particular to academic spoken discourse, it is indicated in this study that the framework to be adjusted and expanded in these aspects in order to be more well-suited for analyzing reporting clauses in academic spoken registers. |