The Covid-19 pandemic has increased media and reader attention to science. Among an unprecedented diversity of different sources, targeted to different audiences, and in the face of the disinformation crisis, traditional media emerge as the most credible source of information on the pandemic, strengthening (or not) their bond of trust with their readers.

Our ambition in this paper is to mobilize a demanding theoretical framework to carry out an international comparison of Covid-19 coverage and reader reactions on legacy news sites in two countries, Brazil and France.

The theoretical framework is an update of Eliseo Veron’s Reading Contract. His theory, originally developed for print media then extended to television, needs adapting to online media communication. Nevertheless, his approach taking into account content and framing but also enunciation modes is fundamental to understanding the bases of media communication, whatever technical form it may take.

Three news outlets are analyzed in each country, during the early pandemic, in April-May 2020. We found some outlets to be analogous in both countries. For instance each has a reference elite news outlet (Folha do Sao Paulo and Le Monde) which adopted very similar modes of addressing the readers, with enunciators familiar with and supportive of science devoting considerable space for science explanation and context in their stories. Popular newspapers (O Dia and Le Parisien), however, covered Covid-19 news quite differently; although their enunciators share a distance from science, often portrayed as a mysterious world, this appears differently in the two journals - besides giving much more space to Covid-19 science news than O Dia, Le Parisien makes contextualization and popularization efforts, offering a bridge for their readers to connect with science, considered relevant.

Reader commentaries vary strongly between the two countries - being much more numerous in French newspapers - but even more so among outlets in each country and among news topics - controversial topics regarding the therapeutic effects of hidroxichloroquine tended to generate a wave of political comments, including conspiracy theories, while less controversial topics tended to generate more technical comments, including questions about the methods of reported research.

We were thus able to demonstrate the pertinence of the Reading Contract model in carrying out a fine-grained comparison between our two countries.