B1 addresses the question of how lexical CG management devices are put to use in non-default CG updates. In the first funding period, discourse particles and focus particles will be the phenomena under investigation. By combining data from experimental and diachronic studies, we will in particular test the hypothesis that (putative) unidirectional diachronic developments in pragmaticalization are fueled by processing advantages. Whether these devices play a role in perspective-taking is a further research question that will be addressed.
