Growing scientific fields often involve multidisciplinary investigations in which the same concepts may have different meanings. Here, we examine the case of ‘gesture’ in comparative research to depict how conceptual diversity hidden by the label ‘gesture’ can lead to consistently divergent interpretations in humans and nonhuman primates. We show that definitions of ‘gesture’ drastically differ regarding the forms of a gesture and the cognitive processes inferred from it, and that these differences emerge from implicit assumptions which have pervasive consequences on the interpretations claimed by researchers. We then demonstrate that implicit assumptions about scientific concepts can be made explicit using a finite set of operational criteria. We argue that developing theoretical definitions systematically associated with operational conceptual boundaries would allow to tackle both the challenges of maintaining high internal coherence within studies and of improving comparability and replicability of scientific results. We thus offer an easy-to-implement conceptual tool that should help ground valid comparisons between studies and serve scientific inquiry.