Community of Fate or Communities of Place? Studying Transnational Discursive Linkages on Twitter during the Covid19 Pandemic

The European public sphere has been under study for more than two decades. During the same period of time, successive political crises have shaken the European Union. They have had inconsistent or even paradoxical effects with respect to the emergence of a transnational public sphere. Recent trends have been discussed using the interwoven developmental terms of Europeanisation and politicisation. By impairing traditional gatekeeping mechanisms and national container logics of industrialised media systems, social media such as Twitter have arguably served these trends by facilitating the formation of more fluid transnational issue publics. This paper takes up some of these strands and studies the temporal and time-varying integration of issue publics on Twitter during the major crisis Europe (and the entire world) is currently going through: the Covid19 pandemic. Do we observe new trends of transnational integration in Europe’s societal debates? Or do we see more nationally structured debates and communicative insulation driven by institutional nationalism in political crisis management? If overarching trends remain inconsistent, which correlations between the course of the pandemic and crisis management on the one hand, and transnational communication on the other can be observed? How do national user communities differ in this regard? In order to study these questions, we rely on Twitter data obtained from the TBCOV database, i.e. a dataset for multilingual, geolocated Covid19-related Twitter communication. We selected corpora for the 27 member states of the EU plus the United Kingdom. We defined three research periods representing different phases of the pandemic, namely April (1st wave), August (interim) and December 2020 (2nd wave) resulting in a final set of corpora comprising 51,893,966 unique tweets for the comparative analysis. In order to measure the level and temporal variation of transnational discursive linkages we conducted a spatiotemporal network analysis of so-called Heterogeneous Information Networks (HINs). HINs allow for the integration of multiple, heterogeneous network entities to better represent the complex discursive structures represented in social media debates. For the initial construction of our HINs, we utilised different sorts of automatically extractable entities, namely hashtags, mentions, named entities, URLs and retweets, that correspond to different dimensions of discursive linkages. For our variable metrics, we counted instances for each of the discursive linkages represented by meta-paths via one of the entity types. Therefrom, we obtained an aggregate measure of transnational linkages on a daily base by relating these linkages back to their geolocated authors. Furthermore, we utilized the Covid Stringency Index, a day-wise composite measure based on indicators like school openness, work related restrictions and travel bans, to check for correlations between national crisis management and (trans-)national communication patterns on Twitter. Preliminary findings suggest that the share of transnational discursive linkages correlates with the severity of the pandemic. Moreover, transnational linkages are higher during the second than during the first wave, indicating effects of institutionalisation. Stringent political measures of crisis management at the domestic level (such as lockdown decisions) seem to provoke stronger national structuration of Covid19-related Twitter discourse.