Peter Hall and Rosemary Taylor's recent article, "Political Science and the Three New Institutional-isms" (published in Political Studies), provides a balanced and meticulous review of the many faces of the "new institutionalism" and a distinctive contribution to the growing literature in this area in its own right. Though deeply sympathetic to the analysis presented by Hall and Taylor, our aim in this response is to draw attention to, and to reflect upon, some of the leey theoretical and conceptual issues that they leave largely unresolved. We suggest that the fundamentally different (and, we argue, profoundly antithetical) ontological assumptions of rationalist and sociological institutionalism mue any attempted synthesis that aims to transcend this divide undesirable and ultimately fruitless. Indeed, we suggest, the ontologies underpinning both rational choice and sociological institutionalism (a calculus and a cultural logic respectively) militate against the development of an institutionalism sensitive to in, stitutional change over time. Consequently, the further development of institutional theory necessitates a distinctive social ontology itself grounded in a clearly articulated view of the complex relationship between structure and agency. Such a social ontology can be discovered in certain of the more generic comments of some historical institutionalists. It is, nonetheless, profoundly at odds with both the "cal· culus" and "cultural" logics which Hall and Taylor claim to identify within the historical institutional ist oeuvre. Accordingly, we reject the temptation to forge a synthetic institutionalism capable of tran scending the limitations of each distinctive perspective, emphasising instead the potential offered by a reinvigorated historical institutionalism that can differentiate itself both analytically and ontologically from rational choice and sociological institutionalism. We outline an alternative interpretation of the social ontology on which this might be premised.
