Folk psychology, cognitive ethology, and artificial agents

In this dissertation I explore two primary questions: (i) is there a scientific methodology for the study of agency in artificial agents such that conclusions drawn from this methodology are compatible with conclusions drawn in human psychology and cognitive ethology, and (ii) under such a methodology, what conclusions can be drawn about modern artificial agents and what comparisons can we make relative to humans and non-human animals?;After introducing a breed of animats I call "pandas", and exploring a social learning experiment conducted with them, I apply the methods of cognitive science to the study of panda behavior. I use these methods to parallel my analysis of panda behavior to the study of animal behavior, in particular, chimpanzee social learning behavior. As these methods have been used to cast doubt on "advanced" social learning in chimpanzees, I use the methods to show that the social learning simulated in my experiment, as a typical artificial life experiment, is very "basic".;I tackle these questions by first reviewing and exploring the standard methodology of cognitive science, as articulated by Daniel Dennett. I confirm that this methodology is properly applied to artificial agents and is capable of facilitating the desired comparisons. This confirmation is made by reviewing evidence for this claim and comparing it to cognitive science work with non-human animals.