Default Reasoning Using Monotonic Logic: A Modest Proposal

Abstract

first order paper presents, a simple extension of to include a.default operator. P rodicatEflog+c iules inference s ecified, overning the operator are and a mo % el sen ences involvin theory for interpretin is develo e s % based on standard 4 default operators arskian semantics. The resu ting system is trivial1 ar sound. It is argued that (a) this logic provi es an adequate basis for default reasoning in A.I. s stems, and (b) unlike most i! t is purpose, retains the lo its proposed for Egic, 1 v rtues of standard first order including both monotonicity and simplicity. Reasoning from incomplete information and from default, generalizations follows P atterns which;s?Fctd f+rst order predicate o ic does not making ' l?he most sttz;;ing inferences devia, ion i;;;lv;t conclusions counterindicated by further information which does nc&explicitly contradict anything previously In standard logics, if a set of premises entails a conclusion, those premises also aneY,tt;;r set containing all that conclusion. Logics with this P roperty are called monotonic. The above departure rom standard logic's reasoning patterns has led researchers to adopt and dev;t;zg non-monotonic logics for use in A.I. systems