This chapter illustrates the results of a workshop held at the University of Salento, during an English-Italian Translation Course. A group of undergraduate students was asked to produce an intralingual translation for the subtitles of the video Capsized in Lampedusa – Fortress Italia, which deals with the situation in Lampedusa after a boat capsized in October 2013. The original video – available on YouTube – is subtitled into English for Vice News and, in a ‘role play’ exercise (Rosnow 1990), students were commissioned to adopt a type of English that could be directed at a wider international audience also including non-native English speakers. The contrastive analysis of the original and reformulated versions is designed to enquire into the influence and actualisation of the commissioner’s requests in the lexical, structural and functional features of target texts. The extent will also be explored, to which the type of English adopted in the alternative subtitles can be defined as a lingua-franca variation (Seidlhofer 2011), due to the inclusion of specific verb tenses and the selection of simplified syntactic structures (Seidlhofer 2004), aimed at facilitating the audience’s reception and accessibility to the semantic dimensions of the text. Finally, the analysis will also highlight the translators’ attempts to respect the temporal and spatial constraints of subtitles (Neves 2009), as well as the multimodal construction (Kress and van Leeuwen 2006) of audiovisual texts, which result in the use of condensation strategies (Gottlieb 2005; Pedersen 2011), thus proposing the new definition of audiovisual mediation.