A knowledge representation language based on integration of such means as frames, semantic networks, production rules, subdefinite computational models, and methods of constraint programming are considered. An important feature of this language is the possibility of operating with objects which can have slots with imprecisely defined (subdefinite) values. Another important feature of the language is that it allows one to bind to any object a set of constraints defined on the values of the object’s slots. These constraints, represented in the form of usual logical expressions, not only serve for control of consistency and correctness of the slots’ values but also allow one to refine automatically subdefinite values. Besides, the possibility of including the operation of implication in constraints makes it possible to provide local inference within the object.