The Integrative Approach to Metadiscourse Revisited

This chapter is going to end where it started - with the discussion of metadiscourse. Applying the ‘integrative approach’ represented by Hyland’s model to two academic genres uncovered several important points. First, the integrative approach enables the researcher to characterize the genre from different viewpoints since it is not limited to text organizing elements or elements referring only to the text. Genres can thus be studied from all the aspects that do not add to the propositional content of the text (even though the term propositional content is controversial itself). Metadiscursive categories are clearly interrelated and the actual expressions in the text are multifunctional (e.g. it is now important to reflect on), signalling both textual and interpersonal meanings. Furthermore, the integrative approach covering different interpersonal aspects seems suitable for the analysis of academic genres since the expressions of interpersonality are not so frequent and expected in comparison with other genres.

On the other hand, the integrative approach has its drawbacks. As already mentioned at the end of Sect. 1, it has been criticized for covering different language phenomena that cannot be easily put under one umbrella term. Hyland defined metadiscourse as “the cover term for the self-reflective6 expressions used to negotiate interactional meanings in a text, assisting the writer (or speaker) to express a viewpoint and engage with readers as members of a particular community” (Hyland 2005, p. 37). Furthermore, he considers non-propositionality to be one of the key principles characterizing metadiscourse, stating that “metadiscourse is distinct from propositional aspects of discourse” (Hyland 2005, p. 38). However, both of these criteria can be challenged.

If we look at all the categories in Hyland’s model, it might be possible to classify them in the following way (see Table 8): endophoric markers, frame markers, code glosses and transitions can be considered self-reflective in a narrow sense in that they refer to the current text. However, they differ in the degree of explicitness of *

Table 8 Hyland’s model of metadiscourse revisited

Self-reflective metadiscourse (textual reflexivity)

Endophoric markers (refer to other parts of the text)

High explicitness


Low explicitness

Frame markers (signal text boundaries)

Code glosses (elaborate propositional meanings)

Transitions (express relations between clauses)

References to other texts

Evidentials (refer to information from other text)

Stance (epistemic and attitudinal)

Hedges (withhold complete commitment to a proposition)

Boosters (allow writers to express certainty in what they say, close down alternatives)

Attitude markers (indicate the writer’s attitude to propositions)

Writer-reader interaction

Self mention (the explicit authorial presence in the text expressed by first-person pronouns)

Engagement markers (explicitly address readers)

The emphasis is mine.

text reflexivity, with endophoric markers being probably the most explicit and transitions the least.[1] Evidentials are not self-reflective but refer to other textual sources, covering citations, paraphrases etc. Hedges, boosters and attitude markers can be considered pragmatic categories commenting on the content of the propositions - they express epistemic and affective stance, thus being self-reflective in a different sense. It is the writer’s evaluation of the state of affairs expressed in the propositions. Finally, self mention and engagement markers are primarily addressing writer-reader interaction.

The non-propositionality as another criterion of metadiscourse also seems problematic. ‘Proposition’ is a semantic term, which originated in logic, and as such it is not easily transferable to discourse analysis. The traditional truth-conditional criteria do not apply here because a number of metadiscursive statements can be described as true or false. However, even if we loosen the criteria and distinguish “things in the world and things in the discourse, propositions and metadiscourse”, as Hyland proposed (Hyland 2005, p. 38), there are still many cases which remain problematic. Considering attitude markers, for example, one of Hyland’s examples reads: “The basis of the enormous productivity and affluence of modern industrial societies is their fantastic store of technological information” (Hyland 2005, p. 164). It is questionable whether “enormous” is an expression of the writer’s attitude. While Hyland interpreted it as an attitude marker, i.e. as non-propositional (thus qualifying as metadiscourse), it could equally be argued that it contributes to the proposition expressed by the text. Similarly, certain hedges are believed to affect the propositional content, e.g. approximators (somewhat, sort of, approximately).

What seems to be a common denominator in all of Hyland’s categories is the writer’s explicit presence in a discourse. Although it is undoubtedly a matter of degree, it can be argued that it is the writer who comments on the form of the text, he or she also expresses stance towards its content and interacts with the reader. Considering the Jakobsonian model of language functions, Adel (2006) regards the metalinguistic function as the indispensable one in her reflexive model. In Hyland’s approach (and generally all broad approaches to metadiscourse) it would be the expressive function which is crucial (although in a more general sense than formulated by Jakobson [1980, p. 82], i.e. “a direct expression of the speaker’s attitude toward what he is speaking about”).

Generally, it seems that the integrative approaches to metadiscourse have moved away from text reflexivity as the capacity of a language to refer to or describe itself, and instead foreground the interpersonal meanings. Whether this approach is justified today (when the interpersonal aspects of language are conceptualized under the headings of stance, evaluation or positioning) is difficult to say. However, at least in academic writing the concept seems to have been useful in showing how writers project themselves into their discourses in order to structure them, negotiate meanings and engage readers as discourse participants.

Acknowledgements The study was supported by the research project CZ.1.07/2.3.00/20.0222, ‘Posflern rozvoje Centra vyzkumu odborneho jazyka anglictiny a nemciny na FF OU’ [Centre for the Research of Professional Language], funded by the Ministry of Education, Youth and Sports of the Czech Republic.

  • [1] Degrees of explicitness in text reflexivity are discussed in Mauranen (1993), who also considersinternal connectors to be of low explicitness.
< Prev   CONTENTS   Source   Next >