How cooperation, institutions, and ethical order arise from long-term interaction remains fragmented across disciplines and is typically explained using psychological or value-based variables. Starting from an irreducible existence condition, I introduce an external alignment structure consisting of a deviation structure d, a consistency reach, and a trust update quantity T. Within this framework, agent interaction and action selection are formalized solely as sequential trust-update structures, without invoking internal states or normative assumptions. Cooperation and related social phenomena are described as observable appearances of update dynamics rather than intrinsic attributes. The theory clarifies how temporal appearances and differential descriptions in classical physics arise as reference constructions, and therefore does not presuppose physical time. In the limit of sufficiently extended consistency reach, self-interested optimization behavior and trust-maintaining behavior become observationally indistinguishable. This scale-invariant construction takes two-agent interaction as the minimal non-degenerate unit and extends uniformly across social and artificial systems. This paper presents a concise English short-form exposition of the Long-term Rationality Principle (LRP), whose complete original formulation is available separately.
Naoaki Tsuda (Tue,) studied this question.