Empowerment-Driven Learning: An Evolutionary and Computational Framework for Academic Motivation

High-achieving students often undergo abrupt, total academic disengagement once terminal credentials (e.g., college acceptance) are secured. This phenomenon—Rational Detachment—reveals a structural gap in the motivation literature: proximate frame-works like Self-Determination Theory describe the psychological conditions under which learning thrives, but none specify the biological mechanism that evaluates whether the physiological cost of cognitive effort is warranted at all. We propose Empowerment-Driven Learning Theory (EDLT) to fill this gap, drawing on evolutionary psychology, Life History Theory, and the computational logic of Active Inference. EDLT models the learner as a metabolic economy: biologically secondary knowledge receives authorization only when its expected return in survival security or social status credibly exceeds the opportunity cost of updating internal predictive models. The framework reconceptualizes motivation not as a psychological state but as a kinematic quantity—the rate of change of expected utility over time. Rational Detachment emerges from this account as an ecologically rational phase transition rather than a character deficit: when the expected utility gradient reaches zero, the system shifts lawfully from Growth into Conservation. Four falsifiable boundary conditions are derived, and a coordinated program of measurement—ecological momentary assessment, life history indexing, and a novel Ancestral Relevance Scale—is proposed to test the framework’s core predictions.