Models of Deliberation in the Social Sciences R. P. LOUI Department ( loui@cs of Computer uastl.edu> Sczence, Was/ungton Unluerslty, St LOUZS, MO AI s extent could never be restricted to its computational modeling of the mind. The idea of a computational model, one that reserves a place for the processing of information that details how constructions may vary with expended resources, is the kind of idea upon which intellectual empires are built. I am referring to recent incursions in the social sciences, specifically in the modeling of legal and economic processes. In the modeling of legal argument, nonmonotonic reasoning has been imported and reformed: the emphasis is on computation and protocol, not syntax and semantics. The models are of debate and argument, dynamic social phenomena, not of mathematical logic pressed and shaped in bizarre proportions. In the social sciences, where social authority is exercised through rules, there is no question of whether defensible rules can exist without a discoverable probabilistic basis. The issues are how rules arise from decisions in prior cases (how case-based reasoning relates to rule-based reasoning) and how the rationales for rules (the reasons for their adoption) affect which counterarguments can be made. We formalists
/lp/association-for-computing-machinery/models-of-deliberation-in-the-social-sciences-aaeNFcddJj