Constraint databases are a recent generalization of relational databases. Constraint databases have greater flexibility and expressive power than relational databases because they allow complete information (unambiguous data) to be described implicitly. The project addresses the development and implementation of new methods for the (1) representation and querying of incomplete information (ambiguous data) in constraint databases, (2) updating and revising single constraint databases, and (3) combining and in case of conflict arbitrating between or among several constraint databases. Approaches for representing incomplete information will be analyzed for time and space efficiency, while update, revision, and arbitration operators will be tested for the least change property as expressed by the Katsuno-Mendelzon and related axioms. A constraint database query and transformation language allowing both representation of incomplete information and update, revision, and arbitration operators will be implemented and tested on applications in spatial databases, telecommunications, and refinement-based modeling of objects. This project will greatly increase the real-life applicability of constraint databases.