The Scarcity Theory of Specialization: Polymathy, the Architecture of Knowledge, and the Question Posed by Artificial Intelligence

Modern science policy, professional training, and the research university share a foundational assumption: excellence requires specialization. This article argues that the assumption rests on a category error. Specialization is a rational adaptation to the cognitive scarcity of individual human minds, not an intrinsic property of high-quality cognition, and a civilization that mistakes the adaptation for the ideal will misread both its intellectual history and its technological present. Drawing on the economics of innovation, studies of creative eminence, and the sociology of science, the article develops a scarcity theory of specialization and tests it against three bodies of evidence: the recombinant structure of major discoveries, the documented liabilities of deep expertise, and the avocational breadth of eminent scientists. The article then disaggregates specialization into three distinct functions, cognitive economy, social coordination, and epistemic certification, and argues that large-scale artificial intelligence dissolves only the first. Foundation models constitute a natural experiment demonstrating that breadth and depth can be jointly scaled when memory and time constraints are relaxed. Four falsifiable propositions and implications for doctoral education and research funding follow. The deeper principle organizing knowledge, the article concludes, is not division but integration.