ABSTRACT
Where Critical Information Theory’s answer to the question of the value of sociological theory in the information age takes the form of social philosophical reflection on the role of information in society, a second set of answers has emerged from within empirical sociology. Following McFarland, Lewis and Goldberg, I refer to this second approach as Forensic Social Science (FSS). The theory that is used within major scientific approaches to sociology is sometimes categorized as ‘empirical theory’ in contrast to social theory or grand theory. Compared to social theory, empirical theory is less abstract; it involves the development and testing of meso theories and theoretical models of narrowly circumscribed objects of study rather than theories of society, social class, capitalism, culture or other broad categories of social phenomena. In these approaches to sociological theory researchers develop, test and refine relatively narrow theories related to specific social phenomena, such as theories of cognitive biases or of gender discrimination in hiring. Fundamental changes in the practice of acquiring and sampling data present major challenges for these approaches to theory. Big data and contemporary computational tools allow for access to data about basic social behaviors that have always been practiced but were previously rarely documented. Such data are often unprompted trace data—records of actual social behaviors rather than information expressed in an artificial relation with a researcher, such as in the context of a social survey or ethnographic interview. Chapter 5 reviews the FSS approach to using sociological theory in big data analysis, evaluating the claims of FSS in light of contemporary trends toward abductive inference in digital social science and the history of inductive inference in social science research. The chapter’s evaluation of FSS is positive overall, although it must be acknowledged that FSS is not a theory per se so much as a paradigm for using theory for empirical research.
