Comunicació presentada a: Speech Prosody 2016, celebrada del 31 de maig al 3 de juny de 2016 a Boston, Estats Units.

For years linguists have noted that intonation patterns encode similar meanings as sentence-final discourse particles across languages. A question that arises is what is the division of labor between the two and whether we find a compensatory distribution between the two kinds of systems. In this article, we focus on two languages within the Romance group (Catalan and Friulian) which have been reported to use intonation and discourse particles to different extents to mark epistemic meanings. Thirty speakers were asked to participate in a Discourse Completion Task designed to elicit statements with several degrees of speaker commitment. The results show that Catalan and Friulian display an asymmetry in the marking of epistemically-biased statements: while Catalan uses a greater variety of stance-marking intonation contours, Friulian uses a more varied set of stance modal particles and a more restricted set of intonation contours. However, both languages make use epistemic adverbs together with intonation and place restrictions on how pragmatic particles and intonation are paired, indicating that the relationship between the two systems can be quite complex. Overall, we claim that dynamic semantic models enable us to integrate the study of intonational meaning with other parts of the grammar into a unified approach.