Secondary Universe Theory: A Proposed Integrative Framework for Narrative Belief Construction, Neurochemical Maintenance, and Resistance to Correction

Why do intelligent individuals construct and defend belief structures that diverge from observable reality, even when they possess the cognitive resources to recognize the divergence? This paper proposes Secondary Universe Theory (SUT), a speculative integrative framework that draws on findings from cognitive psychology, reward neuroscience, social psychology, and terror management research to explain the construction, maintenance, and persistence of what the author terms secondary universes—shared narrative structures characterized by clearly defined functions and systematically undefined mechanisms. The model is offered as a set of theoretically grounded, falsifiable hypotheses rather than a validated causal account. The model proposes a six-stage sequential process, each stage individually supported by peer-reviewed evidence but the full chain currently unverified as a unified causal sequence: (1) default agent attribution as a low-cost cognitive shortcut to which the system reverts whenever mechanism evaluation is structurally unavailable—developmentally (immature PFC), situationally (time pressure or cognitive load), or neurochemically (cortisol-induced suppression or acute LC-NE-induced disconnection); (2) hypothesized stress-induced sensitization of the mesolimbic dopamine pathway via cortisol; (3) proposed hijacking of the anticipatory reward system, in which the nucleus accumbens responds more strongly to anticipated than to actual reward; (4) replacement of reality-contact by imagination, in which vivid mental simulation activates reward circuits while depleting motivation for goal-directed effort; (5) establishment of a cortisol–dopamine feedback loop that renders the secondary universe the dominant and most readily available source of neurochemical relief under conditions of chronic stress and limited alternative reinforcement; and (6) perseverative resistance to correction driven by emotional flooding that blocks prefrontal encoding of alternative strategies. The model explicitly grounds its central construct—the deus ex machina as a constant function—in a tripartite neurochemical mechanism integrating the teleological bias literature (Kelemen, 1999), stress-PFC research (Arnsten, 2009, 2015), dopaminergic prediction error findings (Ongchoco et al., 2023), and pupillometric evidence (Mækelæ et al., 2024). The paper reviews evidence for each stage, positions the model relative to five major existing frameworks and four adjacent research traditions, identifies specific explanatory gaps each leaves open, and proposes a structural diagnostic as a domain-general identification tool. A dedicated section derives neurochemically-grounded predictions about exit conditions, including the counterintuitive prediction that disengagement is most likely during periods of relative safety rather than at the point of maximum crisis. The paper concludes with twelve falsifiable predictions, proposed experimental designs, a measurement appendix, and an articulation of the model’s limitations.