Interpolation of Benchmark Problems in Defeasible Reasoning

摘要

Recently, the research program of defeasible reasoning seems to be driven by aset of simple case studies, called benchmark problems, used to show up the strength (or weakness) of particular modes of defeat. Illustrating that such a pragmatic approach is not entirely satisfactory, the author suggests a counterpart of the interpolation theorem in mathematical analysis. The interpolation theorem for defeasible reasoning will be formulated in precise terms, and will be proven formally. As a result, the positive statement that a particular formalism 'captures all the important benchmark problems' is losing much of its informative content.