When Interpreters Really Are Machines: The Use of Artificial Intelligence in Intercultural Health Care Communication

Machine translation (MT) and artificial intelligence (AI) are used in an ad hoc manner to negotiate intercultural communication in health care settings. As AI is currently subject to public debate, it is important to consider perceptions of and discourses around it. Through a literature review of translation and interpreting studies and health sciences, this article first explores some challenges that MT and AI pose, along with the persistence of reductive thinking around language and the role of the interpreter. It then develops a qualitative content analysis of 41 semi-structured interviews with health care and intercultural communication actors in Catalonia. Informants view MT and AI as useful for simple, low-stakes communication while preferring humans. Despite this, there is an ‘acquiescent anxiety’ around the unstoppable progress of science. In addition, there is a tendency to distinguish between ‘mere’ translation and other ‘added value’ activities, like intercultural mediation. We argue that, when interpreters are not merely thought of as machines, but really are machines, the relentless reduction of interpreting to ‘mechanical parroting’ paves the way for its automatisation. On the flipside, informants reclaim an expanded interpreter role that involves knowledge of language, culture, medicine, the sociology of health and illness, social science methodologies and ethical intercultural communication.