This paper reconsiders the possibility of applying the procedure of falsification, which consists in testing a theory by confronting hypotheses derived from the theory with empirical data, in the studies of culture, in particular when evaluating interpretive hypotheses. Falsification, to which, according to Popper and his followers, the natural sciences owe their success, is viewed with strong suspicion when the object of investigation is meanings and values rather than material phenomena. If by interpretation one understands reconstruction of the artefact’s meaning, obvious challenges when falsifying interpretive hypotheses include: the multiplicity, instability and ambiguity (indefiniteness or self-contradictory character) of meanings inherent in artefacts. All of this does not seem to exclude the possibility of identifying as misreadings interpretive hypotheses which clearly contradict relevant (non-contradictory) artefactual evidence. Falsification thus understood seems indispensable in educational contexts. At the same time, it must be admitted that in practice (as contrasted with the logical formula which underlies the procedure) falsification is inconclusive, and the application of the procedure is further complicated when meanings and values are the object of research. This is one of the reasons why falsification in the humanities (as elsewhere) needs to be complemented by other epistemic procedures while the status of literary studies or history of philosophy might best be perceived as partly cognitive (and in so far as the research conducted in these disciplines is empirically testable, also scientific) and partly creative (artistic).