Governance of Common Knowledge in Distributed Systems

We address a persistent, under-theorised problem at the intersection of distributed systems and governance: how to maintain trustworthy, tamper-evident shared state across parties who may not fully trust one another, not merely at deployment, but across the full lifecycle of a long-lived sociotechnical system. We call this the Common- Knowledge Problem (CKP ) and define it precisely: CKP is the challenge of maintaining knowledge that would be knowable in principle—if authorised by policy and enabled by communication availability—where both the authorisation gate and the channel are governed by whichever entity or rule-based protocol a community has chosen to trust. We argue that the core difficulty is not cryptographic but political : the adversarial forces that degrade governance over time are precisely those analysed by 2,500 years of political philosophy. We propose a layered ontological model that decomposes any CKP system into five analytically separable strata, and maps eight canonical governance archetypes— drawn from Hobbes, Plato, Aristotle, Polybius, Rousseau,Weber, Michels, and Schmitt—onto that stack. For each archetype we identify its principle of legitimacy, its characteristic decay vector, and the conditions under which it constitutes a correct engineering choice. We further derive eight cross-cutting decay laws and a six-step diagnostic procedure for practitioners. Our synthesis is empirically grounded: we test each predicted decay path against documented modern cases in blockchain governance, certificate-authority politics, DAO voter concentration, and enterprise change management. The framework yields a practical method for selecting, instrumenting, and periodically re-diagnosing a governance structure suited to a given threat model and responsible-ownership context.