Cognitive Stable States as the Ontogenetic Foundation of Causal Explanation: With a Commentary on Salmon's Blue Chalk Counterexample

Why do distinct attributes bear unequal explanatory weights in causal explanation? Though both momentum and colour follow conserved‑quantity transmission in physical interactions, momentum is universally recognised as a core causal factor while colour is dismissed as a trivial epiphenomenon. This long‑standing puzzle, formalised as Salmon’s blue-chalk counterexample, has plagued the philosophy of science for nearly three decades. I argue that this persistent paradox stems not from defects in existing causal mechanism theories, but from a cross‑level cognitive misalignment. Treating human cognition as a resource‑bounded information compression process, I identify four naturally emergent discrete cognitive stable states: primordial xiang (L1), symbolised xiang (L2), formalised xiang (L3), and algorithmic xiang (L4). I demonstrate that Salmon attempted to resolve an ontogenetic puzzle rooted at the primordial xiang tier using formal and algorithmic tools that belong to the formalised–algorithmic xiang tiers. The primitive bias for explanatory relevance arises from evolution‑shaped embodied filtering—not from physical conservation laws or statistical correlation. I further formulate six fundamental relations between cognitive stable states: Compression (P1), Generation (P2), Realisation (P3), Dependence (P4), Emergence (P5), and Rollback (P6). Compression and rollback form the complementary dual wings of cognitive dynamics. As an intrinsic compensatory mechanism for the unavoidable information loss incurred by compression, cognitive rollback dismantles rigid stable configurations and restores cognitive freedom—rather than merely reversing compression. Together, the six propositions constitute a unified ontogenetic framework of cognition, which provides a foundational grounding for the theory of causal explanation. Moreover, this framework demarcates the structural boundaries and upper capacity limits of artificial causal reasoning in AI systems.