This special issue offers a critical dialogue around the myriad political dimensions of Big Data. We begin by recognising that the technological objects of Big Data are unprecedented in the speed, scope and scale of their computation and knowledge production. This critical dialogue is grounded in an equal recognition of continuities around Big Data's social, cultural, and political economic dimensions. Big Data, then, is political in the same way in which identity, the body, gender, sexuality, race and ethnicity are political, that is, as sites of struggle over meaning, interpretations, and categorisations of lived experience. Big Data is political in the way circuits of production, distribution, and consumption are political; that is, as sites where access, control and agency are unequally distributed through asymmetrical power relations, including relations of data production. Big Data is political in the way contemporary politics are being reshaped by data analysis in electoral campaign strategy, and through state surveillance as strikingly evidenced by the Snowden revelations on the NSA and GCHQ. Big Data is also political in the contestation of this advanced scientific practice, wherein the generation of data at unprecedented scale promises a precise and objective measure of everyday life. However, the computational dreams of an N = all verisimilitude -that is, of datasets providing a one-to-one correspondence to a given phenomenon -are haunted by the normative biases embedded in all data. This is not to suggest that Big Data -more specifically processes of datafication 1 -are best or at all understood as socially constructed. Indeed, discursive analysis or unreconstructed social theory cannot fully grasp how data re-articulates the social, cultural, political and economic in a deeply recursive manner. Thus, any political reckoning must equally account for the materiality of data, alongside the logic guiding its processes and the practices that deploy its tools. In short, what are the power relations animating the knowledge generated by data analytics? No politics, just data?As smartphones proliferated last decade, powerful computational media were diffused across time and space into a distributed networks of pervasive data generation. An apolitical vision of Big Data quickly followed. Prominent researchers seized on the potential of this data to fuel new forms of computational social science or what some termed 'social engineering' (Lazier et al. 2009). Enthusiasm for developing rigorous mathematical models and applications to understand and predict complex social phenomena reached a high point with Alex Pentland (2014). His MIT research lab developed a highly anodyne vision of "reality mining" (Eagle/Pentland 2006) our data-driven society, wherein the sheer deluge of data points would help attenuate previous limitations imposed by partial or incomplete samples. Reality mining, as proposed by Pentland, looks for social patterns in the quotidian data we generate to infer our relationships, significant locat...