Decision Under Structural Constraint: Feasible Action Before Preference (v1.1)

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Version 1.1 update — This revision adds bounded worked examples and reader-accessibility improvements to the September 27, 2026 v1.0 publication. It also adds an explicit disclosure of AI assistance in drafting, mathematical formulation/manipulation, computational checking, adversarial review, and research organization, together with clarification of the paper’s independent-research and peer-review status. The revision does not modify ASD/BAS v2.9 canon, establish new empirical validation, or alter the paper’s principal technical claims. Decision models often begin with a set of available actions and ask which action should be selected. Applied Semantic Dynamics / Branch Admissibility System (ASD/BAS) and its upstream semantic architecture expose an earlier problem: which actions are actually members of the target-relevant feasible set once permission, scope, current executability, resident structural conditions, exit structure, history, and topology are respected? This paper therefore begins before optimization. It distinguishes represented actions from authorized, executable, structurally feasible, and target-reachable actions. It also distinguishes uncertainty about outcomes from unresolved membership in the feasible set. A probability distribution over consequences does not, by itself, resolve whether the action belongs in the target-relevant feasible set; missing authority or executability information can require preserved UNRESOLVED, while mutually incompatible target-live authority or executability information can require preserved CONFLICT, rather than a probabilistic guess. [1–4] Optimization and preference remain compatible with this architecture. They operate after a feasible set has been compiled for the declared target. The contribution is not a universal normative decision theory. It is the structural claim that coherent ranking cannot repair an incorrectly constructed action space.