The Mark of Cognition After Control: Reconstructive Addressability and the Boundary Between Adaptive Regulation and Cognition

Abstract

This paper develops a proposed operational boundary between adaptive regulation and cognition within the broader theoretical architecture of Persistence Geometry. It challenges the assumption that learning, memory, anticipation, goal-directedness, prediction, or flexible behaviour necessarily establish cognition. Instead, it reverses the traditional explanatory approach: granting persistence and control every capability they can causally support before asking what additional function cognition must perform. The proposed distinction is reconstructive addressability: minimal cognition occurs when absent or non-actual relational structure becomes reconstructively available to present control and causally modifies regulation. This criterion distinguishes historical influence from cognitive reconstruction, preserved information from functional access, and prospective control from the ability to recover and use relations not supplied by current coupling. The framework is representation-neutral and substrate-independent, permitting cognitive realization across biological, artificial, distributed, and environmentally scaffolded systems without presupposing consciousness, language, centralized processing, or symbolic representation. It further distinguishes cognitive occurrence from reconstruction accuracy, adaptive success, system identity, and dependency structure. Crucially, the paper establishes an explicitly falsifiable research programme. Four operational conditions—absence or decoupling, relational specificity, causal efficacy, and independent update—provide a basis for experimentally distinguishing reconstructive cognition from sophisticated non-cognitive control. Proposed investigations include selective ablation, carrier substitution, relational corruption, outcome revaluation, and comparisons against increasingly capable prospective-control architectures. The central contribution is not an assertion about which organisms or artificial systems possess cognition, but a testable proposal for identifying the organizational transition at which cognition becomes explanatorily necessary. Cognition does not replace control; it extends control by making absent relational structure available to present regulation.