The impetus behind this seminar series was a study (Writing the Future) funded by the Qatar National Research Foundation and conducted collaboratively by the University of Exeter and Qatar University. The study involved a cross-linguistic corpus analysis of metadiscourse usage in first language (L1) Arabic university students’ argumentative texts in English and Arabic in a university in Qatar, paralleled by ‘writing conversation’ interviews with a sub-sample of the student writers to explore their metalinguistic understanding of metadiscourse used in their own Arabic and English texts. Thus, it explored, firstly, the linguistic differences in metadiscourse usage in argument writing in Arabic (L1) and English (L2), and secondly, students’ metalinguistic understanding of metadiscourse in argument texts. One important finding from the study was that students had very little metalinguistic understanding of the metadiscourse they did use, or of other metadiscoursal features that they could use: indeed, they often discussed the metadiscourse they used without reference to how it was 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, Reference Hyland2005, p. 37). Although the students had a strong understanding of the conventional features of argument writing, principally derived from writing instruction, they had limited metalinguistic understanding of the textual choices they could make to negotiate the relationships between writer, reader and text.

Given what might be thought of as an obvious connection between what writers do in a text and their authorial understanding of the choices they make, it is perhaps surprising that current research on metadiscourse and metalinguistic understanding for writing exist as very separate fields of enquiry with very little interaction between the two. Methodologically, metadiscourse tends to be explored through corpus studies, but metalinguistic understanding is largely researched through qualitative methods, reflecting the fact that metadiscourse has a textual focus, whilst metalinguistic understanding for writing has a writer focus. Theoretically, metadiscourse is a linguistic theory, whereas metalinguistic understanding, as thinking and linguistic decision-making in writing, is predominantly from cognitive psychology. However, to fully understand how learners become proficient writers, it is critical to attend to both the writer and the written text and to benefit from insights from linguistics and cognitive psychology. This seminar series created the opportunity to bring together international researchers in these two fields to explore synergies between the two concepts and to consider more helpful ways to advance both future research and pedagogical practice with a more coherent model of their inter-relationship.

The seminar series comprised three seminars held in May and June 2022: the first and third were half-day online seminars, whilst the second seminar was a whole day, hosted face-to-face at the University of Exeter and synchronously online. The participants represented research perspectives from 11 countries, and was a salient reminder that the COVID-forced virtual meetings are often a valuable way for research to be more inclusive and to give voice to a greater range of international contributions.

The initial seminar set out to map the territory of recent research on metadiscourse and metalinguistic understanding of writing. Through an open call, six presenters from four different countries shared their research, addressing: the role of teacher-guided talk in facilitating metalinguistic understanding (Newman & Watson, Reference Newman and Watson2020); socio-pragmatic dimensions of language use in Italian and British-English speakers (Bartali, ongoing doctoral study); metalanguage and metadiscourse in negotiating new understanding (Mauranen, Reference Mauranen2023); the linguistic/metalinguistic distinction in writing (Batalha et al., Reference Batalha, Lobo, Estrela and Bragança2021); understanding decision-making in annotating metadiscourse in learner English essays (McCallum, current research); and how students use stance and judgment in writing in different disciplines (Lancaster, Reference Lancaster2016).

The purpose of the second full day seminar was to build on this initial mapping of the territory in more depth through creating opportunities for discussion (conducted both face-to-face and using the online interaction tool, Padlet) about the possibilities and potentials of bringing together research on metadiscourse and metalinguistic understanding for writing. The stimulus for discussion was generated by three keynote speakers, each approaching the theme from different angles. The opening keynote, by Ken Hyland (University of East Anglia, UK), reflected on the ‘meta’ prefix in metadiscourse, and how it has sometimes been interpreted as referring to technical terminology to describe language, rather than as the ways writers monitor their ongoing text to make it coherent, relevant and persuasive to a particular community. The keynote offered a brief overview of metadiscourse (see also, Hyland et al., Reference Hyland, Wang and Jiang2022) as the language we use to help others interpret, evaluate and react to propositional information in ways that we intend, and concluded by offering a reflection on the potentialities of considering the ‘meta’ prefix as a means of characterising target discourses for pedagogical purposes. In this framing, metadiscourse offers insights into a communicative context and the perceptions of its participants that can be productive in writing instruction. The second keynote, by Honglin Chen (University of Wollongong, Australia), also explored the notion of ‘meta’, drawing on Halliday's social semiotic view of meaning-making, and focusing on metalinguistic understanding of writing, defined as the capacity to reflect and articulate the making of writing choices (Chen & Myhill, Reference Chen and Myhill2016). She argued that getting ‘meta’ in writing involves more than incorporating explicit discourse markers of textual orientation: rather, it entails a purposeful awareness of meaning-making choices at the sentence, clause and whole text levels. The presentation suggests that the capacity to construe and orchestrate deeper meaning relations or ‘meta-relations’ is a crucial language resource for becoming competent writers. Michaela Mahlberg (University of Birmingham, UK) gave the final keynote and turned the spotlight onto the possibilities offered by corpus linguistic approaches for classroom application. Drawing on data from the Birmingham Stories Corpus – a corpus of short stories written by children and young adults – and on corpus research on Dickens and other nineteenth-century authors (see also Mahlberg & Wiegand, Reference Mahlberg, Wiegand, O'Keeffe and McCarthy2022), she demonstrated how the connection between fiction and the real world is crucial for guiding readers through narratives. She argued that corpus linguistic research on fiction texts might be relevant to the teaching of creative writing, particularly to navigating the relationship with the reader in narrative.

The final seminar considered the possibilities for a Special Issue on integrating research on metadiscourse and metalinguistic understanding, and in doing so, explored the inter-relationships between metadiscourse and metalinguistic understanding of writing. The reflections of both Hyland and Chen in their keynotes on the ‘meta’ prefix is one obvious connection between metadiscourse and metalinguistic understanding in writing. The Merriam-Webster dictionary has a discussion piece on the changing use of ‘meta’ from prefix to adjective (Merriam-Webster, 2019) and notes the commonality of the prefix use in modern terms such as metaphysics, metatheory and metafiction. Drawing on its Greek etymology – meaning ‘among’, ‘beyond’, ‘after’ – the entry maintains that ‘meta- describes a subject in a way that transcends its original limits, considering the subject itself as an object of reflection’. This works well for the concept of metalinguistic understanding, which is commonly defined as reflection on language and its use (Gombert, Reference Gombert1992), or ‘language as the object of observation and the referent of discourse’ (Camps & Milian, Reference Camps and Milian1999, p. 6). Metadiscourse is internal to the text itself, standing beyond the propositional content of the text to comment on or signal the writer's position within the text. It is less concerned with reflection, however, than with rhetorical purpose, both in terms of ‘the writer's management of the information flow to steer readers through a text’ and ‘authorial interventions which personally engage with the content and readers’ (Hyland et al., Reference Hyland, Wang and Jiang2022, p. 2). What may unite metadiscourse and metalinguistic understanding is moving beyond the propositional content of the text to commenting on the text, either internally as with metadiscourse, or externally as with metalinguistic understanding.

A further strand of discussion related to the way metadiscourse offers a structured way to think about how the reader-writer relationship may be shaped through textual decisions. Hyland and Tse (Reference Hyland and Tse2004) maintain that metadiscourse is ‘the range of devices writers use to explicitly organize their texts, engage readers, and signal their attitudes to both their material and their audience’ (p. 156), thus flagging that it is writer-oriented, text-oriented and reader-oriented. But if metadiscourse is used without explicit understanding of how it functions to construct these reader–writer relationships, writers have limited agency to make active textual decisions because they lack the conceptual tools ‘that support students in reflection on the composing process’ (Chen & Vale, Reference Chen, Vale, Chen, Myhill and Lewis2020, p. 148). Pedagogically, developing learners’ metalinguistic understanding of metadiscoursal elements and how they function could be a valuable way of enabling reader awareness and authorial stance. At the same time, participants also noted that this way of integrating metadiscourse and metalinguistic understanding is founded principally on the dynamics of the writer and the text in addressing an imagined reader. It is also important to consider how the reader understands text, and not to assume all readers are the same. There is very little research that has examined reader responses to metadiscourse in texts or to writers’ efforts to ‘manage’ their responses.