Communicative Effectivenes in Multimodal and Multilingual Dialogues

Multilingual communication enabled by a multimodal speech-to-speech translation system may differ from "ordinary" monolingual conversation int he conversational structure and in the way gestures are integrated in speech. We describe the second of two user studies conducted within the NESPOLE! project investigating these issues. NESPOLE! exploited a client-server architecture to allow an English, French or German-speaking user, while browsing through the web pages of a service provider on the Internet, to connect to an Italian-speaking human agent. Speech-to-speech translation (STST) is provided so that both speakers can use their own native language