Graded membership is a concept in fuzzy logic and cognitive science that allows for partial membership in categories, reflecting real-world situations where boundaries are not always clear-cut. This approach contrasts with binary logic, providing a more nuanced representation of how entities can belong to multiple categories to varying degrees.