Theory and practice of software development: Stages in a debate

Abstract

Starting from the experience gained in organizing TAPSOFT'85, the paper discusses the place of formal methods in software development. It distinguishes two notions of theory: the mathematical science of computation and the treatment of computing as a human activity. An adequate software theory needs to take both theoretical perspectives into account. Therefore, the paper explores the borderline of formalization and human activity in several directions: concerning the role and scope of formalized procedures, the relation between formal models and situated use, the process of learning in software development and the ways computer programs become effective in use. Fundamental assumptions underlying formal methods and their relation to emancipatory approaches such as participatory design are discussed. The paper closes with calling for a dialogical framework for further pursuing these questions.