The author sets out to analyse the linguistic and pragmatic factors involved in the mode-shift from soundtrack speech to written subtitle. His major goal is to find out what happens to communication when speech is represented in writing in a necessarily truncated form, with the added constraints for the reader of discontinuity between one utterance and the next and short processing times. More specifically he examines the loss that occurs in subtitling between two languages and cultures, mainly by looking into the nature of this loss and its potential effects.