Mechanisms play a crucial role in scientific research across various disciplines, and philosophers of science have devoted significant effort into understanding their ontology and epistemology. This paper examines the relationship between mechanisms and phenomena, highlighting the inherent dependence of mechanistic delineation on the characterization of phenomena. By acknowledging that characterizing phenomena is influenced by pragmatic considerations and research interests, the paper argues that mechanistic inquiry is inherently shaped by researchers’ perspectives. This dependence raises concerns about the possibility of a realist (...) view of mechanisms. To address these concerns, the paper explores how top-down constraints, rooted in researchers’ interests and pragmatic concerns, can be balanced by bottom-up constraints derived from empirical considerations. In conclusion, I argue that the interplay between these constraints forms an empirical and realist counterweight to the perspectival nature of top-down constraints. (shrink)