Support Routing and Macroscopic Branch Separation Are Dissociable in a Contextually Modulated Balanced Adaptive Map

This record presents a computational study of finite-amplitude transition dynamics in a deterministic balanced adaptive map with context-dependent higher-order modulation. The work focuses on how edge-support patterns become organized before a macroscopic separation event, and on whether the mechanism that selects this support pattern is distinct from the mechanism that determines whether the separation subsequently occurs. The study synthesizes a sequence of prospectively frozen experiments together with explicitly labeled development-only and negative-result branches. The central intervention tests examine (i) whether preserving the identity of shared contextual “witness” nodes matters for future support organization, and (ii) whether a support pattern that has already become persistently aligned with the future exact split target can remain intact when the subsequent selector dynamics are simplified. The results support a model-specific separation between support routing and later macroscopic branch separation. Contextual same-witness alignment contributes causally to which edge support becomes persistently organized, while a distinct post-lock intervention can alter subsequent split behavior without substantially erasing that established support alignment. Additional experiments characterize local boundary geometry, reciprocal/contextual precursors of support organization, post-lock consolidation, and a later low-gain projective bias whose minimal mechanism remains unresolved. This deposit includes the manuscript, source files, frozen experiment records, machine-readable results, provenance logs, execution manifests, selected development and negative-result records, experiment registries, SHA-256 integrity manifests, and reproduction material. Historical provenance gaps and reconstructed runtime lineage are disclosed explicitly rather than silently normalized. The claims are restricted to the tested deterministic computational model and protocols. The work does not establish a universal law of transition dynamics, a physical spacetime mechanism, cosmological structure formation, or any direct mapping from the model variables to physical observables.