In tandem con l'IA

This contribution draws on an empirical study carried out within the Master’s degree programme in Language Sciences and Cultural Studies for Special Needs at the University of Parma, in collaboration with the University of Mainz. Its aim is to reflect on the use of AI for text revision and for the informal learning of German as a second language. The participating students took part in two tandem cycles: one with a human peer (in chat) and the other with AI, both designed to elicit corrections of short written texts on a given topic and to receive explanations of the corrections. The results emerging from the empirical study will be analysed on the basis of both subjective (1) and objective (2) parameters: (1) The feedback questionnaires completed by participants at the end of the tandem will provide insight into the perceived learning experience and language-learning motivation; (2) A detailed comparison between ‘human’ and ‘virtual’ corrections will allow us to assess the linguistic accuracy of the revised texts, the explanatory value of the feedback, and above all the linguistic aspects (lexis, syntax, etc.) on which peers and AI tend to focus most. Finally, the findings of the empirical investigation will be situated within the existing scholarly debate on the use of artificial intelligence in the field of language learning and text revision.