The problem of the attractor: A singular generality between sciences and social theory

Contemporary complexity sciences claim a literal, non-metaphorical applicability to physical, economic, social and cultural events. They envision the development of a general social or historical physics. Conversely, in the social sciences and humanities, complexity sciences have been typically treated as a source of new metaphors or tropes to be used in theory-building. Can there be a critical social or historical physics that is not a world-view and that does not treat science as a source of metaphors? The Lorenz attractor figures centrally in the history of complexity science as a popular image of 'deterministic chaos' and the 'butterfly effect', as an indication of how far complexity science has progressed in the last two decades, and, as this article argues, as an event whose multiplicity of interpretations attests to the problem it raises, the problem of generality associated with complexity. Via the Lorenz attractor, the article examines three attempts to treat complexity non-metaphorically in recent theoretical work (Delanda; Massumi; Stengers). In these accounts, the attractor performs several different functions. It forms part of a re-engineered concept of multiplicity, it helps conceptualize feeling or sensitivity, and it raises the general problem of practice in theory-building.