Abstract
Pencil and paper are perhaps the most effective problem-solving tools ever invented. Why is this so, and what does this portend for computer-assisted problem solving? In this paper we investigate why the computer has not made more significant inroads into many aspects of problem solving, even in domains ostensibly concerned with purely formal methods. We observe that for many problem-solving activities computers are currently more constraining than enabling, particularly during problem formulation. We identify some of the obstacles that must be overcome in making the computer a more attractive medium for problem solving, and explore some of the tools that will likely play a part in bringing this about.
