A major theme in artificial intelligence research has been the quest for methods for common-sense reasoning about actions and their effects. Such methods, when implemented as computer programs, are called automatic problem solvers, and are the subject of this papers.