Functional Cognitive Continuity

This preprint introduces Functional Cognitive Continuity (CCF) as a latent construct for characterizing the longitudinal organization of artificial agents and human–AI systems. The proposal distinguishes momentary task performance from continuity of trajectory. It investigates how intelligent systems preserve relevant memory, decisional coherence, error-correction capacity, provenance, traceability, and intelligibility across successive transformations, interactions, and internal state changes. Δ-Coherence is presented as a constrained vector-based operational assessment framework rather than as a measure of artificial consciousness. The model distinguishes four analytical levels: Architecture; Cognitive Organization; Phenomenology; Observable Behavior. Its empirical scope is restricted to the inference and evaluation of cognitive organization from longitudinal behavioral evidence. The framework does not infer phenomenal experience directly from system outputs. To prevent critical weaknesses from being concealed through aggregation, the proposed evaluation architecture combines two complementary layers: Non-compensatory floor constraints for dimensions considered essential to functional continuity; Multidimensional profiles that can be compared through Pareto dominance rather than reduced prematurely to a single composite score. The manuscript also distinguishes agent-level CCF from symbiotic CCF, which concerns continuity emerging within coupled human–AI systems. It formalizes three admissibility regimes—robustly admissible, fragilely admissible, and non-admissible—and presents an initial matrix of observable dimensions for longitudinal evaluation. The theoretical framework is accompanied by six falsifiable hypotheses and an experimental protocol including baseline conditions, ablation studies, construct-validation procedures, longitudinal measurements, and reproducibility criteria. This manuscript is a theoretical and methodological proposal in a pre-experimental stage. It does not make claims regarding phenomenal consciousness, legal personality, personhood, or the moral status of artificial systems. Its purpose is to provide a testable framework for evaluating whether an artificial or symbiotic cognitive system preserves functionally relevant organization across time and transformation.