Value and Goal as Boundary-Relevance Ranking: Causal Boundary Gradients, Goal Stability, and Value Drift under Finite Capacity

Official website: distinctiontheory.orgPublic portal for the start guide, papers, claim status, failure registry, prior-art boundary, and citation resources. Canonical GitHub repository:https://github.com/yiningwu-research/Distinction-Theory Value and Goal as Boundary-Relevance Ranking is the FDS-M2 working paper in the Distinction Theory / Active Finite Distinction Systems series. The paper develops the value-goal layer of the agency-semantics spine. Building on the formal core of Active Finite Distinction Systems, the M0 agency-semantics paper, and the M1 model of attention as distinction admission, it treats value as FDS-value: causal boundary-gradient relevance under finite capacity, rather than moral value, subjective preference, reward, or intrinsic worth. An evaluand is FDS-valuable for a system only relative to a specified boundary, loss function, intervention grammar, horizon, action or update channel, and resource budget. Its gross causal boundary gain is the finite-difference reduction in expected future boundary-maintenance loss; its net FDS-value subtracts scalarized evaluation, verification, action, maintenance, latency, or opportunity cost; and its risk-weighted value adds collapse-risk reduction near critical thresholds. A goal is defined as a stabilized FDS-value ranking coupled to a policy orientation across finite update windows. The paper distinguishes salience, attention, reward, preference, FDS-value, goal, commitment, and moral value. It introduces predictive-causal dissociation, intervention grammar, gross/net value accounting, evaluation deficit, risk-weighted ranking, goal-stability indices, proxy-boundary divergence, value drift, multi-goal Pareto conflict, collective-goal alignment, ranking synchronization demand, and goal hysteresis. The paper does not derive ethics from boundary maintenance and does not claim that all value reduces to survival. It provides a conservative finite-system bridge for analyzing value and goal as boundary-relevance ranking under capacity, verification, and resource constraints. The accompanying replication package includes deterministic synthetic normal-form code, figures, and CSV outputs. The simulations illustrate predictive-causal dissociation, risk-dominant ranking, goal stabilization, proxy reward hacking, second-order evaluation deficit, Pareto conflict, collective ranking synchronization, and recovery lag. These figures are synthetic normal-form illustrations, not empirical evidence. Scope. FDS-M2 supports later work on meaning as value-preserving semantic compression, trust as delegated evaluation and verification, AI alignment as proxy-boundary auditing, epistemic pollution as collective value-ranking distortion, institutional collapse as value drift or goal rigidity, scientific method as predictive-causal correction, and civilization memory as stabilization of value-relevant records.