If you have a big family, remembering who is in what relation to whom may be a daunting task. Language users might face a similar kind of problem. Decades of work in linguistics show that a small number of structural relations govern a wide range of syntactic dependencies. In particular, the logical scope relation, known in linguistic circles as c-command, has particularly pervasive effects. People seem to be very sensitive to violations of c-command requirements, which suggests that they are able of representing and using relational information. However, it is not clear how exactly they manage to do it: memory models which are currently used in psycholinguistic research don't seem to provide a straightforward way of encoding the c-command relations.

This tension could be addressed by either challenging the assumptions of the memory models (maybe there is some way to accurately represent the relevant information), or by showing that people sometimes are not capable of accurately using the information about syntactic relations (maybe the memory limitations are real and relational information is not perfectly represented/used after all). We investigate the latter possibility. I will present an eye-tracking study in which we explore people's sensitivity to relational constraints by looking at the processing of reflexive pronouns resolution.