The Locus of Legitimate Interpretation in Big Data Sciences: Lessons from -omic biology and high-energy physics

Abstract

© The Author(s) 2018. Over the past decade, ‘big data’ has been positioned as the indispensable mode of 21st century research across academia (Boyd and Crawford 2012; Kitchin 2014i). While many of the foundational concepts and techniques of the big data sciences were already well-established practices across a number of scientific disciplines, only recently have they been assembled into a distinct field of research claiming legitimacy in and of itself (Kitchin 2014i, 2014ii, Ruppert 2015, Beer 2016, Williams et al. 2017). While social science has a quantitative history with ‘big’ datasets dating back to before Durkheim (1897 [2006]), the emergence of ‘big data’ and computationally-intensive social science is a contemporary phenomenon. As with much of the discourse surrounding big data across the board, there is a tendency to posit the application of ‘big data’ approaches to social science questions as a revolutionary innovation in the profession, both in terms of empirical reach and in theoretical advancement.