The current study compared the frequency of Discourse Markers (DMs) based on Fraser’s (2006) semantic model as well as the pragmatic functions of DMs based on Brinton’s (1996, 2017) model in two university textbooks. The first textbook Research Methods in Language Learning was written by David Nunan (a native speaker of English), and the other book Research Methods in Applied Linguistics was written by Hossein Farhady (a non-native speaker of English). The corpus of this study consisted of 14847 words from Nunan’s textbook, and 17706 words from Farhady’s textbook. It was chosen from chapters with similar and comparable content, and Flesch’s (1948) readability formula did not reveal a significant difference between the textual difficulty of the two corpora. Altogether, both corpora contained 772 DMs, in other words 300 DMs in Nunan’s textbook and 472 DMs in Farhady’s textbook. The results revealed no statistically significant difference between the frequency of DMs in Nunan’s textbook and Farhady’s textbook. The range of frequencies from the highest to the lowest belonged to Elaborative, Contrastive, Inferential, and Temporal markers respectively in both textbooks. Moreover, the pragmatic functions of DMs were evaluated by two other raters besides the researcher, and inter-rater reliability showed a high correlation (0.99). There was not a significant difference between the frequencies for functions of DMs either. The textual functions found in both textbooks included opening markers, closing markers, topic switchers, information indicators, and sequence markers. However, no function was observed at the interpersonal level. Possible pedagogical implications have also been mentioned.
