Testing pragmatic accounts of complex linguistic constraints

If a speaker says “John said that he was tired”, “he” probably means “John”. But if a speaker says “He said that John was tired”, “he” cannot mean “John”.

In this work package we are using behavioural experiments and computational modelling to build and test a pragmatic account which explains both why languages have these types of restrictions and how children come to learn them.

Project Team: Ben Ambridge (lead), Liam BlythingSilke Brandt, Anna Theakston

Duration: 3 years, starting March 2021

Project Number: 2.1