This study has explored DMs conceived as signals of semantic relations between adjacent or more distant segments of discourse and their role in the enhancing of an interpretation coherent with the author’s communicative intentions, thus contributing to both cohesion and coherence. Since these language phenomena are mostly used intentionally by writers as guiding signals in order to enable the prospective reader(s) to arrive at an intended interpretation of discourse, it has been considered important to discover whether there is any variation, both cross-culturally, i.e. English vs. Czech expert writers, and in terms of experience/expertise, i.e. Czech expert vs. novice writers, in academic written discourse, represented in this chapter by RAs on the one hand and Master’s theses on the other.
Based on the findings drawn from the analysis of causal and contrastive DMs used in order to build coherence relations in academic discourse, it can now be postulated that a certain cross-cultural variation has been identified. This variation is caused in particular by the fact that RAs produced by native speakers of English resemble nonacademic texts in some respects, for example, in the extent of interactivity between the writer and his/her potential reader(s), which is reflected, among other things, in a more frequent use of certain DMs typically connected with spoken discourse, such as informal conversation; this concerns, for example, the markers because, so and then from the group of causal DMs and but from that of contrastive DMs. As for the most common markers within their respective semantic classes, there is not much variation, as exemplified above. As regards the differences between expert and novice academic writing, it can now be concluded that there are certain noticeable differences, mostly related to the redundant use of a rather limited repertoire of certain markers in the learner corpus, in particular causal DMs (e.g. as and because), which Czech students (4.83) apply in many more cases per 1,000 words than English expert writers (3.94), for example. There is also variation in the repertoire of DMs the individual authors included in the study tend to use. This variation results from the writing habits of individual authors, instructions provided by teachers, and/or even topics which may sometimes require a more frequent expression of cause and/or contrast. In addition, with both causal and contrastive relations novice writers unambiguously give preference to the application of paratactic markers, which enable natural ordering and usually quicker processing of discourse segments.
Finally, it should be conceded that only further research in a wide range of different genres of academic discourse, including academic texts from other discourse communities, will allow for more generalizations.