Concepts in Social Cognition: What can we learn about bias from the normative dimension of conceptual representation? (with Kevin Reuter)     (working draft here)

 Recent studies suggest that social role concepts such as those expressed by weldernurse and scientist are unique in that they encode a normative dimension of conceptual representation. This normative dimension can be picked out by expressions such as true scientist, and its properties can be investigated by examining the acceptability patterns of those expressions under various manipulations of background information. In this paper, we report three studies which show that gender has a significant effect on the acceptability of true-modifications of social role terms. The studies show that this effect cannot be reduced to a simple mis/match between the gender of the individual engaged in a social role and the gender associated with that role. We argue that the observed effect is the result of a particular gender bias such that arbitrary members are assumed to be more committed to their social roles if they are male compared to female. We call this bias the "role-commitment gender bias". The studies show that this bias is a default which operates when there is no background information about the level of commitment of members to their roles, and can be cancelled once that information is explicitly provided. We examine in detail how these results bear on questions such as how biases like the role-commitment gender bias are encoded in certain socially relevant concepts, what kinds of interactions trigger their application, andto what extent the role-commitment gender bias behaves (in response to information) like a propositionally structured belief.