Overview and publication identity

Discrete rules that individuate relational or joint-dependence boundaries can change discontinuously under arbitrarily small perturbations of an underlying stochastic process. We exhibit this failure exactly in a finite joint-output model, derive a continuous reconstruction-based diagnostic using directed statistical deficiency, and determine when the resulting effective relational descriptions compose under causal and resource constraints. We then apply these results to the SPC-2 localization constitution as a motivating case. In the masking family, an arbitrarily small singleton effect deletes a projected joint-dependence edge while substantial joint information and the contrast of a fixed return schedule survive. The exact deficiency optimum quantifies what is lost by deleting part of the observation and is invariant under common-conditional padding. We separate joint information, executable native return, and resistance to resource-constrained cut simulation; the resulting schedule-indexed candidate families need not be partitions. For finite marked stochastic instruments, strong quotients admit substitution in matching causal contexts and minimization-compatible composition, while transfer of cut radii additionally requires control of the permitted simulator-law families. Finally, an attribute is identifiable from exact complete experiment laws precisely when it factors through their operational quotient. Functional premises leaving a local phenomenal predicate unconstrained do not determine it; SPC-2 supplies an explicit connecting premise. The results identify a robustness vulnerability and constrain successor doctrines without deriving a unique admission law, validating phenomenal presence, or identifying source ontology from records alone.