From Risk to Polyuncertainty: Rethinking Systemic Risk Governance in an Uncertain World

Contemporary systemic-risk governance is organized around a family of concepts, risk, Knightian uncertainty, deep uncertainty, complexity, systemic risk, and polycrisis, each built to isolate one dimension of a disorderly world. We argue that the family has a gap at its center: none of its members names the standing condition in which multiple uncertainties interact and evolve continuously, including in the quiet periods before any recognizable crisis. We call that condition Polyuncertainty and define it through five necessary attributes, multiplicity, interdependence, dynamic evolution, cascading effects, and limited governability, and five generative mechanisms. The contribution is deliberately conceptual. We propose no new estimator and claim no priority over the mature apparatus for systemic risk in finance or compound risk in climate science; the claim is that a bounded organizing concept, positioned against its neighbors and equipped with boundary conditions and testable propositions, can improve problem definition and set a cumulative research agenda across governance capability, measurement, and domain application. Five recent episodes in which losses spread through structure rather than the initiating hazard motivate the concept, and a transparent agent-based simulation, offered as a proof of concept, shows that the five mechanisms generate a disproportionate-cascade signature single-hazard assessment misses, and that slack, coordination, and the timing of coordination move it as the framework predicts.