Fuzzy logic extends classical bivalent reasoning by allowing propositions to take continuous truth-values between 0 and 1, thereby modelling graded uncertainty and vagueness in a formal setting.