FIRST PERSON REPRESENTATIONS NEED A METHODOLOGY BASED ON SIMULATION OR THEORY

Although their thesis is generally sound, Barresi & Moore give insuffi cient attention to the need for a methodology, whether simulation-base d or theory-based, for choosing among alternative possible matches of first person and third person information. This choice must be sensiti ve to contextual information, including past behavior. Moreover, apart from simulation or theory, first person information would not help pr edict future behavior.