Dulcinati, G. & Pouscoulous, N.
Division of Psychology and Language Sciences, University College London
According to Grice (1989), implicatures are afforded by the hearer?s assumption that the speaker is cooperative. We used a novel experimental paradigm to manipulate this assumption. Participants played an internet-based game with a confederate who was either cooperating or playing against them. Participants listened to the confederate?s descriptions of grids, whose squares may or may not contain stars, and they had to choose the grid that was being described from a set of three (or indicate that they did not know). Some descriptions (e.g. here the squares in the bottom row have stars) could give rise to Ad Hoc quantity implicatures (i.e. no other squares have stars). In Experiment 1, participants in the cooperative condition chose the grid consistent with the implicature significantly more than participants in the uncooperative condition. This result is consistent with Grice?s account. In Experiment 2 we used Ad Hoc and scalar implicatures. We found that our manipulation had the same as in Experiment 1 for Ad Hoc implicatures whereas it had the opposite effect for scalar implicatures (result not significant by subjects). We discuss the unexpected trend of scalar implicatures in relation to the findings of Doran et al. (2012).