COVID-19: Constructing a Specialised Sample Corpus to Test Trends in Deontic Obligation

This paper reports on the construction and analysis of a specialised sample corpus made of corpuslinguistics, deontic modalitytranscripts from official UK Government statements from the COVID-19 daily briefings between March and April 2020. It tests trends in deontic obligation proposed by Leech et al. (2009) by applying a re-defined parameter of source through a qualitative analysis of competing modal forms in the data. The forms under analysis are the ‘core’ modals ‘must’ and ‘should’ and the semi-modals ‘have to’ and ‘need to’. Source (i.e. the object at the origin of the obligation) is operationalised following Depraetere and Verhulst (2008) and distinguishes between discourse internal/external and intersubjective (mixed) sources of necessity, where discourse external is divided between circumstance, condition and rule/regulation. Results from the analysis show that the data conforms to the trends suggested by Leech et al. (2009), both in terms of competing modal frequency counts and their differing functions. ‘Need to’ is the most frequent deontic modal in the corpus, used predominantly with discourse-internal sources which highlights the collective nature of its use. ‘Have to’ has largely regulatory discourse-external sources, used by the Government to refer back to pre-existing regulations. When the obligation arrives for the first time via a discourse-internal source (i.e. when the Government issue new rules), ‘should’ is overwhelmingly favoured, particularly in negative constructions with the second person singular i.e ‘you should not’. This reflects its categorisation as a ‘less face-threatening’ alternative to ‘must’, which is consequently the least frequent modal in the corpus. Taken together, it seems that ‘should’, ‘have to’ and ‘need to’ are indeed beneficiaries of ‘must’s decline in contemporary use. It further suggests that the Government deem a more collective and less directive strategy of obligation to be more effective when seeking the conformity and compliance of the general public, reflecting perhaps a less admirable attitude towards strong authoritative instruction within society. The study also shows how applying the redefined parameter of ‘source’ can prove insightful to the subtly differing functions of competing modal forms in spoken language.