Abstract.According to the normative theory of legal science, juristic ought‐sentences describe rules, since legal science just deals with rules, and rules cannot be described but by means of ought‐sentences. The author challenges this view. Two different constructions of “describing rules” are proposed: Namely, either interpreting or stating the validity of rules. “Interpreting rules,” in its turn, can be understood in three different senses: listing all the possible meanings of rule‐formulations, reporting the different interpretations a rule‐formulation has in fact received by (...) courts, or ascribing meaning to rule‐formulations. However, the author argues that ought‐sentences are not the proper tools to accomplish such tasks. At the same time, juristic ought‐sentences cannot be understood as validity statements, since they neither mention any rule whatsoever, nor include the term “valid.” Further, if validity‐statements were ought‐sentences, their logical behaviour would reflect the logic of rules themselves. However, as the late Kelsen argued, things do not run this way, since two inconsistent ought‐sentences, if understood as validity statements, can paradoxically both be true, as well as both false. Hence, validity‐statements cannot be reduced to ought‐sentences iterating the rules which they (are supposed to) refer to. (shrink)