Computational Social Science of Social Cohesion and Polarization

Abstract The emergence of Computational Social Science (CSS) marks a transformative shift in the study of social phenomena, triggered by advancements in computational technology and large-scale collection of (digital) behavioral trace data. CSS promises to provide an accurate and deep understanding of social behavior through the analysis of social interactions, movement patterns, and communication networks at scale. However, the lack of a common theoretical language and clearly defined scientific agenda complicates CSS’ ability to build a common understanding of social phenomena. In this chapter, we establish CSS as a scientific paradigm, emphasizing the integration of computational tools with rigorous theoretical modeling. Specifically, we position CSS within the trajectory of a maturing science and emphasize the synthesis of data-driven CSS with theoretically-informed modeling for understanding social phenomena like social cohesion and polarization. To illustrate how CSS can advance these fields specifically, we unpack a range of conceptualizations of political polarization and social cohesion, and demonstrate how they have evolved in Europe over recent decades. Finally, we discuss how the subsequent chapters in this book aim to bridge the gap between theoretical and empirical CSS for social cohesion and polarization.