Research of formulaic language in academic writing has primarily investigated the use of single types of formulaic sequences in academic research articles in various disciplines. Studies in this line of research have revealed dramatic variations in the use of formulaic language across academic disciplines (e.g., Cortes, 2004; Hyland, 2008a; Jalali & Moini, 2014; Shahriari, 2017). However, there is evidence that discipline alone does not tell the whole story about linguistic variation (Gray, 2015). Different varieties of texts within one discipline may reflect different linguistic characteristics depending on specific communicative purposes (Biber & Conrad, 2009). It follows that the almost exclusive focus on the academic research article may “limit our knowledge of the discourse practices within discipline” (Gray, 2015, p. 19). Moreover, formulaic language encompasses different types of sequences (e.g., collocations, lexical bundles, frames, etc.) each of which only reveals a partial picture of formulaicity in discourse (Wray, 2005). Thus, studies that investigate the use of single types of formulaic sequences may provide only partial descriptions of the registers they investigate. Therefore, to better serve disciplinary writing instruction, there is a need for studies that provide more comprehensive descriptions of formulaic language in various registers within one discipline.

The present dissertation takes a step in that direction by investigating within-discipline linguistic variation through the comparison of the formulaic profiles of two registers in the field of medicine: the medical research article (MRA) and the medical case report (MCR). These two registers that have both been reported in the medical literature to contribute to advancing research, clinical practice, and education in the field (e.g., Man et al., 2004; Rison et al., 2017). The study proposes a more comprehensive approach to the description of formulaic language and investigates the use of various formulaic sequences that have been described as accounting for the formulaicity of discourse. Such sequences include: (a) collocations, pairs of words that tend to co-occur, (b) multiword collocations, sequences of three or more words with strong mutual attraction (such sequences consist primarily of lexical words, most of which are technical terms), (c) lexical bundles, most frequent sequences of three or more words in a register, described as the building blocks of academic writing (Cortes, 2013), and (d) frames, sequences of three or more items with one variable slot. Frames have been described as allowing writers to make more creative use of formulaic language (e.g., Biber, 2009; Gray & Biber, 2013).

The analyses of the formulaic sequences in the two registers often revealed structural similarities but noticeable variations in terms of the discourse functions of the sequences. Such variations reflect the differences in the situational characteristics of the two registers such as communicative purposes, nature of data and evidence, textual organization, to name but a few. The findings of the present study portray MRAs and MCRs as two distinct registers, thus highlighting the importance of investing within-discipline variations to better serve disciplinary writing instruction.