In recent years, much has been written on big data' in both the popular and academic press. After the hubristic declaration of the end of theory' more nuanced arguments have emerged, suggesting that increasingly pervasive data collection and quantification may have significant implications for the social sciences, even if the social, scientific, political, and economic agendas behind big data are less new than they are often portrayed. Compared to the boosterish tone of much of its press, academic critiques of big data have been relatively muted, often focusing on the continued importance of more traditional forms of domain knowledge and expertise. Indeed, many academic responses to big data enthusiastically celebrate the availability of new data sources and the potential for new insights and perspectives they may enable. Undermining many of these critiques is a lack of attention to the role of technology in society, particularly with respect to the labor process, the continued extension of labor relations into previously private times and places, and the commoditization of more and more aspects of everyday life. In this article, we parse a variety of big data definitions to argue that it is only when individual datums by the million, billion, or more are linked together algorithmically that big data' emerges as a commodity. Such decisions do not occur in a vacuum but as part of an asymmetric power relationship in which individuals are dispossessed of the data they generate in their day-to-day lives. We argue that the asymmetry of this data capture process is a means of capitalist accumulation by dispossession' that colonizes and commodifies everyday life in ways previously impossible. Situating the promises of big data' within the utopian imaginaries of digital frontierism, we suggest processes of data colonialism are actually unfolding behind these utopic promises. Amid private corporate and academic excitement over new forms of data analysis and visualization, situating big data as a form of capitalist expropriation and dispossession stresses the urgent need for critical, theoretical understandings of data and society.
