The study of decision making has been dominated by economic perspectives, which model people as rational agents who carefully weigh costs and benefits and try to maximize the utility of every choice, without consideration of issues such as cultural norms, religious beliefs and moral rules. However, psychological findings indicate that in many situations people are not rational decision makers as defined by the economic theories. One of the domains in which traditional cost-benefit models fail to predict human behavior is the domain of moral reasoning.

This work presents the first computational model of recognition-based moral decision making, MoralDM, which integrates several AI techniques in order to model recent psychological findings on moral decision making. MoralDM uses a natural language system to produce formal representations from psychological stimuli, reducing tailorability. The impacts of secular versus sacred values are modeled via qualitative reasoning, using an order of magnitude representation. MoralDM uses a combination of first-principles reasoning and analogical reasoning to model the recognition-based mode of decision making.

The results of MoralDM experiments provided the impetus to further examine the role of cultural narratives and analogical reasoning on moral decision making. This work examines whether the processes by which core cultural narratives are applied in people's lives follow the principles of analogical retrieval and mapping. In particular, it examines how analogical accessibility and alignability influence the use of canonical moral narratives. The results of a series of experiments performed among Iranian and American participants are reported and these results are simulated using MoralDM.

The last contribution of this thesis is regarding the use of structured qualitative representations and analogical generalization in modeling the similarities and differences in causal reasoning for biological kinds. The individual models built from transcript data are used to construct generalizations, which are tested both by inspection and by creating a classifier to distinguish models based on the culture and the level of expertise of the participants.

Overall, this thesis argues for the importance of highly structural representations in conjunction with analogical reasoning for capturing and modeling some effects of culture on cognition.