More distinguishes three uses of logic in AI; as a tool of analysis, as a basis for knowledge representation, and as a programming language. A large part of the effort of developing limited-objective reasoning systems goes into the management of large, complex bodies of declarative information.