“So and so” says, states and argues: a corpus-assisted engagement analysis of reporting verbs

Cassi L. Liardét*, Sharyn Black

*Corresponding author for this work

    Research output: Contribution to journalArticlepeer-review

    14 Citations (Scopus)


    A key feature of academic texts is the heteroglossic interaction that occurs between the writer and the experts he/she references. When integrating outside experts into their texts, writers often employ integral, author prominent reporting structures, selecting reporting verbs (RVs) to evaluate the veracity and merit of the propositions. This paper examines EAL and English L1 learners’ RV use and compares it with that of experts, providing a corpus-assisted, comparative analysis. It further explores how writers build intertextuality through these RV choices. Using the resources of APPRAISAL theory, and specifically, the ENGAGEMENT system, we found that experts tend to favor dialogically contracting RVs (e.g., show and find) that endorse the proposition whereas learners rely heavily on expanding RVs that entertain the evidence as an option to consider (e.g., suggest) or simply attribute it to an outside expert (e.g., state). In particular, both the EAL and English L1 learners strongly rely upon more “neutral” attribute: acknowledge structures (e.g., state, according to), providing no overt indication as to their intersubjective stance on the evidence. These comparative findings provide a roadmap for novice writers to develop authorial stance and adapt to the expert conventions of their given fields.

    Original languageEnglish
    Pages (from-to)37-50
    Number of pages14
    JournalJournal of Second Language Writing
    Publication statusPublished - Jun 2019


    • reporting verbs
    • intertextuality
    • referencing
    • stance
    • engagement
    • appraisal theory
    • academic discourse


    Dive into the research topics of '“So and so” says, states and argues: a corpus-assisted engagement analysis of reporting verbs'. Together they form a unique fingerprint.

    Cite this