Polygenic Prediction: How Contested Genetic Technologies Organize Fields of Expertise

Abstract Prediction has long been a feature of medical genetics. Yet with the advent of polygenic indices (PGIs), there has been a fundamental shift in the complexity and ambiguity of the predictive power afforded by genetics. These indices, which use large datasets to estimate how genetic variants relate to phenotype, are in principle calculable for any medical condition or social behavior: cancer, depression, educational attainment, or even height. Yet PGIs are a contested technology, characterized by uncertainty that originates in scientific institutions and is amplified through application to clinical, commercial, and policy settings. The paper examines how this technology organizes scattered forms of expertise – including population genetics, clinical medicine, public policy, and psychiatry – into a nascent scientific field. Using empirical examples, we demonstrate how actors inhabit social spaces with locally specific forms of symbolic capital that are connected by a common technology. This technological field ’s specificity comes from the way it is layered on top of other fields, thereby defining legitimate uses of PGIs, as well as the way it facilitates the interpellation of other forms of capital into the field. We argue that this could be a productive framework for future sociological research on topics including artificial intelligence and causal inference.