Making Social Science Matter

In the AI Story

The book draws on Hubert Dreyfus's critique of artificial intelligence from the 1970s — a connection that becomes operationally prescient in the AI context two decades later. Dreyfus argued that expert human judgment is not reducible to rules, and therefore cannot be replicated by rule-following systems. Flyvbjerg extended the argument from the individual expert to the social world, showing that social phenomena are constitutively irreducible to the kind of regularities that context-independent knowledge requires.

The argument's methodological consequence is radical. If social phenomena are context-dependent, then the dominant methodology of modern social science — designed to produce context-independent, generalizable findings — is structurally incapable of producing the most important form of knowledge about its own subject matter. The instruments produce findings that hold across populations but tell you nothing about what to do in the particular situation you face. The social sciences have been measuring the wrong thing with the wrong instruments, not through incompetence but through inheritance of an inappropriate framework.

The alternative Flyvbjerg proposes is phronetic social science — a research program organized around the production of practical wisdom rather than universal laws. The methodology is case-based, longitudinal, context-sensitive, and value-laden. It takes particular situations seriously as objects of inquiry; it follows them over time horizons adequate to the processes being studied; it treats context as the essential dimension of the phenomenon rather than noise to be controlled for; and it engages explicitly with questions of power, value, and judgment that positivist methodology excludes.

The book's reception was mixed and revealing. Quantitative social scientists received it as an attack on their methodological commitments; qualitative researchers embraced it as philosophical vindication. The response demonstrated the very institutional divisions the book diagnosed — a research paradigm organized around the epistemic aspiration was institutionally structured to dismiss the phronetic alternative. The AI transition has since exposed the limits of the epistemic paradigm with a force that philosophical argument alone could not generate, bringing renewed attention to Flyvbjerg's framework.

Origin

Flyvbjerg developed the argument through the 1990s, drawing on his Aalborg case study and his engagement with the philosophical traditions of phenomenology, hermeneutics, and American pragmatism. The book was published by Cambridge University Press in 2001 and has been translated into more than a dozen languages.

Key Ideas

Epistemic failure of social science. The century-long attempt to produce universal laws of human behavior has not succeeded because the phenomena do not exhibit the regularities the method requires.

Context as constitutive. For social phenomena, context is not noise to be controlled for but the essential dimension of the phenomenon itself.

Phronesis as alternative. The book proposes practical wisdom — knowledge of how to act in particular situations — as the proper end of social inquiry.

Dreyfus connection. The argument draws on Dreyfus's critique of AI and extends it from individual expertise to the social world, producing a framework whose AI-relevance became operational two decades later.

Methodological program. Case-based, longitudinal, value-laden inquiry replaces the quantitative extraction of universal regularities as the proper form of social research.