“…For example, Hyland (2005b) examined interactional metadiscourse in 240 RAs from eight disciplines straddling the hard/soft division and reported that RAs from the soft disciplines used interactional metadiscourse more frequently than RAs from the hard disciplines. Differences between hard and soft disciplines were also reported for specific types of interactional resources: hedges (Abdi, 2002;Hyland, 1998Hyland, , 2005bVold, 2006), boosters (Hyland, 1998(Hyland, , 2005bPeacock, 2006), attitude markers (Abdi, 2002;Giltrow, 2005), selfmentions (Harwood, 2005a;Hyland, 2001b;Lafuente-Millán, 2010), and engagement markers (Hyland, 2001a(Hyland, , 2002b(Hyland, , 2005b. Taken together, the findings from these cross-disciplinary studies point to disciplinary variation in the use of interactional metadiscourse.…”