Code to fit the RSA-SvI model to data from Leffel et al (2019) on the "not very ADJ" construction.
See A Rational Speech-Act model for the pragmatic use of vague terms in natural language in the Proceedings of CogSci 2022 for theoretical background and details on the model.
The file RSA-SvI-main.Rmd contains all the code besides the Stan models, and RSA-SvI-main.pdf is the knitted output.
I have saved draws from the large Stan models to avoid having to fit the models again, but they are too big for github (1Gb total, about 36Mb per file). Contact me if you want them.