Abstract
We develop two related Actor-Network Theory (ANT) arguments for organizational analysis. The first concerns research strategy and draws upon Latour's (1999) notion of definitional `sliding' to describe how ANT overcomes its analytical limitations by removing conditions that exclude the `other'. Through this discussion, we argue that, research-wise, ANT appears to be ontologically relativist, in permitting the world to be organized differentially, yet empirically realist in providing `theory-laden' descriptions of organization. Our second argument concerns institutional boundedness and flexibility, and suggests that ANT's ontological slipperiness may actually be of value for studies of organizational form. We outline how, under ANT, the analytical focus shifts from structural prescription to processual deconstruction, the associated political dimension concerning where and for whom boundaries are produced/consumed. Overall, we argue for organizational field research that avoids any obligation to impose and defend its own theoretical discriminations.