The language of the COVID-19 Pandemic: Investigating official communication and its relations with collective and individual emotions

The coronavirus outbreak represents an unprecedented global health crisis as the pandemic continues spread, driving massive restrictions on fundamental, everyday-life behaviors (Paules, Marston, & Fauci, 2020). Individuals living in countries around the world have been instructed to profoundly alter their daily routines and behaviors, particularly those involving social contact. A global pandemic like the one we are currently experiencing constitutes not only a crisis of public health, but also major, collective, psychological upheaval that evokes strong emotional responses that require intra- and interpersonal regulation at both the individual and societal levels (Maercker & Horn, 2013). Past research has established that the manner in which disease and the health behavior regimens are communicated can either undermine or facilitate the success of public influence and its subsequent impact (Rubin, Potts, & Michie, 2010; Vaughan & Tinker, 2009). As such, the language itself that is deployed by official health channels plays a central role in modern public health systems, a point that has been increasingly acknowledged in recent years (Gostin, 2018). Moreover, in the current crisis, press conferences by non-medical experts, such as political leaders, have the potential for huge impacts on public health. For example, Chancellor Merkel`s speech on March 18, 2020 had around 25 Millions of viewers; the Swiss Bundesrat's and Boris Johnson's press conferences saw similarly high rates of viewership. Qualitatively speaking, it is evident that each individual government's leaders presented the same essential content - disease information, public advice, behavioral restrictions - in remarkably different ways. However, the complex interplay between each source's linguistic style and the social psychological landscape of each nation remains poorly understood. Our proposed research aims at not only understanding the linguistic fingerprint of these public communications and their differential impact on the general public, but the a priori public receptivity to each fingerprint.Collective emotional responses and coping strategies surrounding crises are directly reflected in everyday language used on social media, such as Twitter (Brummette & Fussell Sisco, 2015; Eriksson, 2016; Garcia & Rime, 2019). Likewise, individual reactions to collective experiences of crises are also reflected in daily language use (Mehl & Pennebaker, 2003). In the context of emerging technologies for the "at scale" automated psychological analysis of language data (Boyd, 2017; Boyd, Pasca, & Conroy-Beam, 2019; Boyd & Pennebaker, 2015b), this international study proposes to investigate both the language use in official webpages and statements during the COVID-19 crisis as well as collective and individual emotions and emotional regulation processes that publicly unfold through Twitter and more targeted, individual writings. Principally, we propose to accomplish through the analysis of language "style" rather than language "content". Importantly, the analysis of language style is a unique approach to psycholinguistic fingerprinting that can be parallelly studied across languages (Meier et al., 2019; Meier et al., 2020; Pennebaker, Boyd, Jordan, & Blackburn, 2015). Our central research questions are: Are there systematic differences between public communications in USA, UK, Germany and Switzerland? More specifically, are there societally impactful divergences in language styles (reflecting, e.g., analytical vs. narrative thinking (Jordan & Pennebaker, 2017; Pennebaker, Chung, Frazee, Lavergne, & Beaver, 2014) and references to the disease and virus ("Corona", "Chinese virus", "COVID-19" etc.)? Are these differences related to collective emotional responses in Twitter over time and to intra- and interpersonal emotional regulation as measured in questionnaires and essays (Horn & Maercker, 2016; Horn, Samson, Debrot, & Perrez, 2019)? Additionally, we will explore how the language in official governmental communications are meaningful contributors to infection rates within their corresponding locales as the ultimate outcome of successful health communication during the crisis. Results of the study will inform further health communication and open the door for automatic tracking of collective emotional responses at scale during a public health crisis in order to allow timely public mental health interventions.