In light of Theory Section Chair John Mohr’s focus this year on Big Data, we the Editors approached scholars from a variety of different subdisciplines to ask for their perspective on how Big Data might shape the future of theory and theorizing in their subdiscipline. In the spirit of Big Data, we asked them to “tweet” their thoughts—though instead of 140 characters, we asked for around 140 words. Ten scholars were gracious enough to respond. Their takes range from celebratory and optimistic to wary and cautioning, but each draws attention to important features of Big Data and how they might interface with theory.
Featuring Julia Adams and Hannah Brueckner on the sociology of knowledge, Adina D. Sterling on organizations and inequality, Elizabeth Popp Berman on education, Neal Caren on social movements, Mara Loveman on comparative historical sociology, Deborah Lupton on health, Robin Wagner-Pacifici on culture, Gary Alan Fine on ethnography, and Juan Pablo Pardo-Guerra on economic sociology and science studies.
Featuring Julia Adams and Hannah Brueckner on the sociology of knowledge, Adina D. Sterling on organizations and inequality, Elizabeth Popp Berman on education, Neal Caren on social movements, Mara Loveman on comparative historical sociology, Deborah Lupton on health, Robin Wagner-Pacifici on culture, Gary Alan Fine on ethnography, and Juan Pablo Pardo-Guerra on economic sociology and science studies.
 Julia Adams Yale University  Hannah Brueckner NYU-Abu Dhabi | #Sociology of KnowledgeBig Data is the new fad in the social sciences and the digital humanities. It's great to finally catch up with the computer scientists, engineers and marketers, right? But since irony is among the many things for which we have no workable algorithm, the latter sentence reveals just one of the enormous challenges for Big Data analysts. And one where theory comes in. Data such as that found on Wikipedia consists of complex texts embedded in human communicative strategies, including power plays, and fissured by systematic absences. To decipher the meaning of such data in a sociological sense, as we have found in our analyses of Wikipedia and academic knowledge production, we must theorize the underlying mechanisms (including algorithms, as it were) that simultaneously produce network structures and holes; signs and regimes; texts and absences. Only then is it possible to differentiate sense from nonsense. |
 Adina D. Sterling Stanford University | #Organizations #InequalityIn the field of organizations and inequality, big data may have a two-fold impact. First, the sheer data available on people and organizations can help scholars uncover new patterns of mobility that were previously obscured, informing our understanding of how opportunities are structured by organizations and society. Second, the use of big data by practitioners – an expanding field called “people analytics” – could shape what scholars find. The fact is the availability of big data varies across demographic groups. A New York Times blog reports there were more CEOs named John than all female CEOs combined in 2015. In analytical terms, this means that statistical models used to predict how a candidate for a CEO position named John will perform will have greater predictive power than for any female candidate. For this and other reasons, it’s important that scholars lead not just in consuming big data, but that we use what we know methodologically to advocate for its responsible use. |
 Elizabeth Popp Berman SUNY-Albany | #EducationBig Data will teach us a lot about education—about how students learn, and about social mobility and reproduction. But it will also have two meta-consequences that deserve our attention. First, it will produce a new wave of overconfidence that we now know how to “fix” education. Second, it will create opportunities for new kinds of organizational actors—collectors and analyzers of data—that will reshape the educational ecosystem and further blur public/private boundaries. With regard to the first, theorists are well-positioned to provide reminders of the limits of social scientific knowledge as a guide to practical action (which is not to say that social science lacks value). The second will require theorists to think through what it means to be public or private, and how to understand educational organizations as outside actors become incorporated into their technical core. |
 Neal Caren University of North Carolina, Chapel Hill | #Social MovementsSocial movement scholars, although not necessarily sociologists, have been at the forefront of analyzing Big Data. Primarily, this is a result of 1) activists and others using Twitter for some visible political purposes, and 2) Twitter allowing researchers access to some data in ways that other social networking sites do not. This has nudged social movement scholars to focus more on contemporary movements and devote theoretical energies to the role of social media in shaping politics. This focus on current events will likely grow as advances in the automated extraction of newspaper data may soon enable a new wave of analyzing protest events. Despite these empirical shifts and their potential to influence new theoretical approaches, recent submission to the social movements journal Mobilization indicate that the dominant theoretical duo of opportunities and frames remains incredibly resilient. |
 Mara Loveman University of California, Berkeley | #Comparative HIstoricalWe might debate whether “Big Data” is the right label for some of the vast new data sources for historical research – things like the exponentially increased number of digitized archival collections around the world, or the release of complete count US Census datasets for multiple years. But there is little question that this transformed datascape dramatically expands both the scope and scale of plausible comparative historical research agendas going forward. New data sources combined with new automated record linkage methods and other innovative computer-assisted analytic techniques open up exciting opportunities to revisit classic problems in comparative historical sociology and to ask new theoretical questions about micro-macro linkages, embedded social processes, and multiple temporalities in explanations of historical continuity and change. |
 Deborah Lupton University of Canberra | #healthSociologists are grappling with how to research the liveliness of the personal data that digital technologies are constantly generating. The world of digital data offers an exciting opportunity to reconceptualize our research methods and theories of social life and sociality. More applied research and theorizing are needed to address the impact of digital data on people's lives and how they interact with, understand and incorporate digital data. I use the term “data sense” to include these dimensions of the sociality of digital data. An interdisciplinary approach has never been more important: science and technology studies, anthropology, and cultural geography, for example, have much to offer sociological concepts of data sense and lively data. For example, sociomaterialist perspectives, as developed in actor-network theory, technoscience, and material anthropology, acknowledge and address the entanglements of humans and technologies. They therefore contribute an approach that goes beyond the discourse-centric position that has tended to pervade critical sociological theorizing in recent decades. |
 Robin Wagner-Pacifici New School for Social Research | #CultureCultural sociology has always been pluralist in both its methods and theories, with one wing developing close readings of cultural objects in the hermeneutic tradition. Such virtuoso interpretations demand a deep knowledge of the objects under examination (be they textual, symbolic, imagistic, or gestural in nature) -- a knowledge from the inside. They move from surface to depth and from text to context in order to ferret out meanings. Recent methodological innovations in computational analysis of texts (including Named Entity Recognition, Syntactic Parsers, and Topic Models) have brought big data into the purview of cultural sociology. These methods identify such things as patterns, co-occurrences, and probabilistic “topics” extant in the data that can elude even the most talented of human close readers. As productive as such tools are, they are severely limited by their very ignorance of the cultural capacities and resonances of the objects under analysis. The challenge is to bridge the virtuoso close reading and the systematic distant reading with a language that incorporates both the hermeneutic insight, with its commitment to singularity, and the positivist findings associated with a more normal science model that aims for validity, reproducibility, and reliability |
 Juan Pablo Pardo-Guerra University of California, San Diego | #Economic Sociology |
 Gary Alan Fine Northwestern University | #EthnographyIn 1995, an educational researcher, Gregory Cizek, penned a cri de coeur entitled, “Crunchy Granola and the Hegemony of the Narrative.” He noted that, in his field, “I’m actually an outcaste [sic] from the academy. I’m a quantitative methodologist. I use numbers.” Fieldworkers, themselves no strangers to disciplinary exclusion, took some wry amusement from his plight. But Cizek had a point. Just as participant-observers became “ethnographers” in an age of the perfect anecdote – the verity of being there - quantitative research needed rebranding. The examination of large data sets is now labeled “Big Data,” and includes measures of internet searches, consumer purchases, or measures of traffic. Because these data are not gathered from individual informants (as in the case of experimental or survey data), big data often involves unobtrusive measures, as pioneered by Donald Campbell. The digital footprints of millions are available to be sorted, measured and mined. These data, seemingly “objective markers” – but of what? -, have become real. They are surely real in themselves, but without motive or motivation, how can we transform to generalization, to prediction, to theory? Despite the allure that bigger numbers are better numbers, Big Data sets aside motivation and motive (accounts of motivation). These studies emphasize the what, the where, and the when, and they leave the how and the why to the sharp vision of the ethnographer or to the big imagination of the data analyst. Because these different methods offer distinctive answers to different types of questions, “big data” and “deep data” need not compete and can, sometimes, jointly nourish. Big Data are not crunchy granola; the question is whether they will serve us as thick or thin gruel. |
RSS Feed
