<p>This paper examines the dilemma of value-ladenness, which suggests that either the social world is inherently value-laden, requiring normative concepts to explain human actions and thoughts (anti-naturalism), or norms lack explanatory role, equating the social with the natural world (naturalism). The first horn implies a methodological gap between the social and natural sciences and risks representing humans as causally idiosyncratic, while the second risks ending up in relativism. My strategy is to treat normative propositions as constitutive principles in empirical research, retaining the explanatory role of norms without presupposing any methodological uniqueness for the social sciences qua social. The contribution of the present paper lies primarily in the philosophy of the social sciences. My argument, in brief, is this: if we take for granted the role of the functional a priori in empirical inquiry, and if we understand norms based on Anscombe’s notion of direction of fit, then the dilemma of value-ladenness can be dissolved.</p>