Being always true by their very form, tautologies should be uninformative; however, they are felicitously used to evoke some sort of shared knowledge. In this paper we explore the varieties of knowledge speakers and hearers can resort to in interpreting nominal tautologies. Our examples show that different dimensions of knowledge (encyclopaedic vs. metalinguistic, normative vs. descriptive, and common vs. local) can combine in different ways in the interpretation of tautologies, and that these combinations fit in well with existing classifications of tautologies in the literature. Tautologies serve an argumentative function by invoking general knowledge as a means to justify a certain behaviour, but they also can have the opposite interpretation: when they are used as replies to information-seeking questions, they can only be interpreted as refusals.