Introduction

In the last 30 years or so, metadiscourse has been used quite frequently as a tool for characterizing various genres, especially the genres of academic discourse. It might be surprising that its popularity endures despite numerous definitions of the concept that vary significantly. It is probably not possible to say any more that metadiscourse is “undertheorized”, as Hyland and Tse suggested in 2004; however, the differences between its conceptualizations are not negligible, but have important methodological and practical implications for research. The common aspect of a number of earlier definitions was their stress on the non-propositionality of metadiscourse. Crismore et al., for example, defined it as “linguistic material, written or

T. Guziurova (*)

Faculty of Arts, Centre for the Research of Professional Language, University of Ostrava, Ostrava, Czech Republic e-mail: This email address is being protected from spam bots, you need Javascript enabled to view it

K. Aijmer, D. Lewis (eds.), Contrastive Analysis of Discourse-pragmatic Aspects of Linguistic Genres, Yearbook of Corpus Linguistics and Pragmatics,

DOI 10.1007/978-3-319-54556-1_10

spoken, which does not add anything to the propositional content” (1993, p. 40). However, this criterion has not proved to be satisfactory, so writers have instead looked for a definition of metadiscourse in Halliday’s theory of metafunctions. Halliday’s ideational metafunction was connected with propositional content; metadiscourse, on the other hand, was believed to convey either interpersonal or textual meanings (Vande Kopple 1985, pp. 84-85).

Basically, it is possible to identify two approaches to metadiscourse^[1] - integrative and non-integrative - depending on whether it includes only text-organizing elements and elements referring to the text itself, or also the writer’s epistemic and affective attitude to the text and interaction with the reader. The integrative (“broad”) approach investigates linguistic elements revealing how the text is organized, but it also focuses on the writer’s presence in the discourse, the ways in which he or she comments on the text or expresses his or her attitudes towards it. The broad approach includes ‘stance’ as a metadiscursive category, usually under headings such as validity markers (hedges, emphatics) and attitude markers. This strand has been applied in many studies, and it was probably taken furthest by Hyland, who suggests that “all metadiscourse is interpersonal in that it takes account of the reader’s knowledge, textual experiences and processing needs” (Hyland 2005, p. 41).

The non-integrative (“narrow”) approach primarily investigates aspects of text organization and elements referring to the text itself. This conception of metadiscourse does not include the writer’s presence in the text in general; rather, metadiscourse is an explicit expression of a writer’s awareness of the current text. Therefore, it is defined as “the writer’s explicit commentary on his or her own ongoing text” (Mauranen 1993, p. 154). The second important feature of the nonintegrative approach is the ‘current text’, meaning that references to other texts are not included within this approach. The narrow approach has been applied (with some modifications) for example by Mauranen (1993), Schiffrin (1980) and Bunton (1999).

Drawing on the integrative approach to metadiscourse, this study focuses on one category, engagement markers, and aims to compare their occurrence and use in two academic genres, the research article and the undergraduate textbook, within one discipline - linguistics. Since the analysis is part of a larger research project aiming at the description of metadiscursive features in the undergraduate textbook, the chapter also briefly discusses the integrative approach to metadiscourse, represented especially by Hyland (2005). Despite its popularity among researchers (e.g. Crismore and Farnsworth 1990; Luukka 1994; Hyland 2005; Bondi 2001; Boggel 2009; Kuhi 2012), the integrative approach has recently been criticized for covering disparate language phenomena that cannot easily be put under one umbrella term (text-organizing elements, the expression of stance, writer-reader interaction).

Therefore, the study also assesses the potential advantages and drawbacks of this model.

[1] The two approaches were probably first distinguished by Mauranen (1993). For a detailed discussion, see, for example, Adel (2006).