The COVID-19 pandemic has highlighted the importance of collecting and disseminating publicly relevant health information from and to lay audiences to combat global health crises. In Flanders, Belgium, reaching linguistic minority groups within this context largely depended on government-mandated contact tracing through telephone interactions, which was initially put in place to contain the level of outbreak, but was gradually tasked with relaying safety measures and other information related to the pandemic. In line with these developments, Flemish contact tracing over the telephone initially supported Belgium's three national languages, viz. Dutch, French and German, as well as English as a lingua franca. This list was gradually expanded with minority languages including, among others, Turkish, Arabic, Bulgarian and Polish. This article reports on a one-year fundamental research project as part of which we recorded and analyzed telephone contact tracing conversations which were conducted in the minority languages Turkish and Arabic. We specifically analyze these Turkish and Arabic data as spaces of linguistic non-understanding from the perspective of the researcher-fieldworker (van Hest, Ella & Marie Jacobs. 2022. Spaces of linguistic non-understanding in linguistic ethnography (and beyond). In "Methodological issues and challenges in researching transculturally," 14-38. Cambridge Scholars Publishing. Available at: https://hdl.handle.net/1854/LU-8772518), and aim to uncover their impact on the interactional dynamics in this 'novel' institutional activity type. Through a Goffmanian analysis, the study reveals how the inclusion of minority languages in contact tracing practice led to complex shifts in participant roles, as contact tracers balanced their institutional responsibilities with imperatives of patient-centered rapport, cultural mediation, and enabling the data collection on behalf of the researchers. The condition of linguistic non-understanding also underlines how the (potential) presence of an overhearing eavesdropper may have influenced the interactional dynamics between contact tracers and index patients. The paper hence provides insights into the complexities of conducting contact tracing in a multilingual context, while simultaneously shedding light on the implications of linguistic diversity on situated institutional practices.