Institutional Nescience: Representation, Correction, and the Limits of Epistemic Jurisdiction

This essay develops a comparative conceptual framework for understanding institutional nescience: a condition in which a representational framework systematically excludes a consequential dimension of reality while institutional mechanisms for recognizing and correcting that exclusion are impaired. The concept is distinguished from ordinary ignorance, specialization, methodological simplification, incomplete evidence, and isolated error by three criteria: systematicity, consequentiality, and epistemic insulation. The central pathology is not representation itself but the transformation of a partial representation into an exclusive jurisdiction over what may count as real, relevant, or knowable. Three cases receive full treatment through a common seven-question schema: colonial enumeration and caste (historical), scientific practice (a contrasting case of institutionalized corrigibility), and large language models (contemporary computational). Cartography, labour commensuration, ecological valuation, and Bhartṛhari's philosophy of language serve as shorter illustrations of the same mechanism. The essay distinguishes first-order from second-order nescience, proposes extensional friction as a system-level condition in which representations remain exposed to independent constraints, and concludes with five requirements of a non-sovereign epistemology: scope transparency, representational non-exclusivity, independent correction, domain-appropriate answerability, and cross-tradition self-critique. The argument is deliberately non-totalizing: marginality does not guarantee truth, representation is not inherently pathological, and the same classificatory operation may protect or obscure depending on its institutional purpose and correction architecture; participatory Indigenous mapping provides a documented example in which representation has supported tenure, resource management, and community claims.