Social theory explanations commonly take one of two forms. Accounts couched in terms of macroscopic entities such as institutions, culture, class, structure and tradition tend to privilege stability and regularity. Individualistic explanations, on the other hand, take these entities to be ultimately reducible to free actions of individuals and are most adept at explaining transformation and volatility in the social realm. These two forms of explanation are rooted in radically different ontological and normative assumptions, and no attempt to connect them has garnered wide acceptance to date. This dissertation once re-examines the tension between them, which has become known as the "structure vs. agency" debate, by drawing on and extending the insights of "theories of practice", a literature that locates the junction of structure and agency in the routines of ordinary daily activities.

The dissertation begins by critically examining two extant theories of practice. One originates in Pierre Bourdieu, Anthony Giddens, and William Sewell's responses to structuralism; the other is articulated by Theodore Schatzki, who draws on Ludwig Wittgenstein to characterize practice as a semantic lens through which social actors make sense of the world. An alternative theory of practice is then developed, inspired by Charles Taylor's Heideggerian conception of embodied agency. This theory of "embodied practice" advances a novel formulation of the structure-agent relationship. Through a fine-grained analysis of the cognitive and informational processes by which practices project a semantic dimension onto the world, embodied practice theory renders robust forms of personal agency compatible with certain forms of semantic macrostructures. The dissertation goes on to describe how embodied practices can account for both change and stability in society and how the concept of an embodied practice may be profitably employed in applied social analysis and political theory.