This essay traces the ontological and political limits of Bruno Latour’s conceptualization of the ‘common world’. Latour formulates this concept in explicating how modernist scientific and political institutions require a metaphysical foundation that is anti-democratic in rigidly partitioning nature from society. In the stead of nature/society, Latour proposes a ‘cosmopolitics’ in which we recognize our embroilment in systems comprised of heterogeneous human and nonhuman actors, and seek to innovate appropriate procedures for governing such systems and composing a more peaceful common (...) world. However, feminist and postcolonialist science studies scholars have argued that Latour’s project fails to apprehend the experiences of subjects marginalized by scientific and political structures of representation. Thus, science studies scholars seem to be faced with a problem; to avoid ontological dissonance and contradiction in integrating Latour’s program with feminist and postcolonialist critiques we must defer his claims that the cosmopolitical program entirely supersedes contemporary social movements and oppositional politics. I offer an alternative. Namely, I argue that Latour’s formulations of cosmopolitics’ procedures and the common world’s boundaries actually require an incipient model of the limits of democratic representation. I modify Latour’s proposed system of representation by reading it against Dipesh Chakrabarty’s concept of ‘subaltern pasts’, which are unassimilable to academic historical narratives (in their attribution of agency to nonhuman gods, for example). This reading emphasizes how Latour does not adequately theorize and politicize the partially autonomous reality of those excluded from the common world and thus fails to attend to how this externalization replicates the violence of modernist representational institutions. (shrink)