Mark Bevir and Rod Rhodes' interpretive political science hardly needs an introduction to policy scholars. Their approach, philosophically sophisticated and grounded in a realistic appraisal of the nature of governing in modern liberal democratic societies, is of considerable significance to anyone in political science, public administration or planning with an interest in interpretive methods. Focusing on the concepts of meaning holism, anti-representationalism, tradition, dilemma and decenteredness, I briefly describe the philosophical foundations and substantive theory of interpretive political science. The main part of the essay consists of a critique of interpretive political science. My argument is that the key concept of practice is underdeveloped. The point of a philosophy of practice is to integrate belief and action to the point where they form one organic activity system for the purpose of moving about effectively in the world. Bevir and Rhodes' insistence on keeping belief and action separate and, in addition, on privileging belief as the major driver of change and adaptation to changing circumstances, inserts a deep inconsistency into their work; an inconsistency that, to my mind, becomes manifest in the relative paucity of their own empirical work. I conclude the essay with a brief exposition of what a practice-based, posthumanist approach to policy research entails (centering on the concepts of agency, the dialectic of resistance and accommodation, temporal emergence and cultural extension) and of what its practical implications are for policy analysis.