Social Spaces Christoph Houman Ellersgaard, Lasse Folke Henriksen, Peter Marcus Kristensen and Anton Grau Larsen The methodological imagination of social spaces This chapter introduces methods for re-imagining international political sociology through social spaces of power, alliances and positions. The spatial methodologies introduced, Social Network Analysis (SNA) and Multiple Correspondence Analysis (MCA), build on a relational ontology that represents the social not in terms of individuals or organizational entities, but connections and interactions, spaces of positions and position-taking. One appeal of spatial methodologies like SNA and MCA, compared to other quantitative methods, is that by ”drawing things together” (Latour 1990) they enable us to visualize hitherto invisible social spaces in graphs and sociograms. As descriptive and explorative methods, they allow for complex relations of interdependence rather than pure one-way relations between independent and dependent forces (Emirbayer 1997). These methodologies are the backbone of the “descriptive assemblages” (Savage 2009) unifying different contemporary sociological traditions. It should be emphasized that spatial methods not only render social spaces visible and thinkable, but also governable. The mapping of social spaces opens some political spaces and closes down, or makes invisible, others. By drawing things together, we also open for doing things together. As such, the methodological imagination of social spaces can be deployed for more or less critical purposes. So far, critical scholars within international political sociology and international relations have largely shied away from these quantitative methods, mostly because they were seen as a defining feature of its constitutive Other: the non-critical, problem-solving positivists who use quantitative methods to reproduce rather than disrupt political order. But criticality is not tied to the methods per se, we argue, which is not to say that the quantitative methods introduced here are apolitical, but, quite the contrary, that they construct political objects that can be deployed for different purposes. Network analyses based on, for example, military alliances, diplomatic connections and co-membership in IGOs construct much more state-centric social spaces than those based on professionals and practices. Similarly, international security is often visualized through organized crime, trafficking and terrorist networks to be policed, but could also, for example, be envisioned as a field of security professionals and bureaucratic apparatuses, which opens a different space of governability. Or take the world of finance as seen through the transnational capitalist network of corporate elites or central bank directors, which represents a vastly different social space, than a network of country level financial flows. Global trade visualized as supply chain networks has different uses than if visualized through deadly logistical and production conditions, as do mappings of terrorist networks for counterinsurgency purposes compared to, say, accounting for terrorism through global inequalities (Aradau and Huysmans 2014:604). To be sure, policy-oriented uses with implications for diplomacy, businesses, intelligence, counter-insurgency and policing are prevalent, and such modes of application risk reifying the object of study and directing the forces of action represented in social networks towards particular pre-defined aims. But the methods themselves allow for a