Non-Causal Coherence Indexing: Studying System Stability Without Diagnosis or Prediction

This paper defines a non-causal research posture for observing system coherence without invoking explanation, diagnosis, prediction, optimisation, or prescription. Coherence is treated as a structural observable rather than an outcome or variable. The framework establishes longitudinal exposure as the temporal condition under which persistence becomes visible and introduces structural admissibility as a precondition for valid observation. Indexing is defined as a non-quantitative marking of persistence or loss under exposure, explicitly separated from scoring, threshold logic, and evaluative classification. The paper delineates strict research design constraints that limit conclusions to observability and prohibit causal inference, comparative assessment, temporal projection, or intervention justification. Non-causal coherence indexing is presented as a bounded methodological framework. It does not propose diagnostic, predictive, or optimisation models and does not merge with adjacent balance, orientation, momentum, or relational constructs within the Atlas series.