The Architecture of Supernatural Belief: Opacity, Satisficing, and the Self-Source

People invent things and sometimes treat those inventions as real, as with supernatural beliefs. This paper asks why minds of a certain architecture do this and whether an artificial mind with the same architecture would. It applies a framework from prior work on emotion, consciousness, and intelligence to the general creative-and-believing capacity, treating supernatural belief as one band on a wider spectrum. The account separates two layers: a wide creative drive produces constructs, and a narrow step holds some constructs as real when felt emotional closure settles a demand and no reach behind the construct finds support.

Study at a glance

Abstract

Minds make things up and treat some of what they make up as real. A person invents a tune and knows it is an invention, yet the same person may hold that there is a god behind the sky and treat that not as a thing made but as a feature of the world. This paper asks why a mind of a certain architecture does this, and whether an artificial mind of the same architecture would. It applies an established framework, developed in companion work on emotion, consciousness, qualia, coherence resolution, and intelligence, to the general creative-and-believing capacity, treating supernatural belief as one architecturally special band on that wider spectrum rather than as the product of a dedicated faculty. The account separates two layers. A wide creative drive produces constructs as the residue of a constrained system resolving what presses on it. A narrow step then holds some constructs as real rather than as the mind's own making. That step turns on two things together: a felt emotional closure that settles a demand, and the absence of any reach behind the construct to find that nothing but that closure supports it. A system that can read how its own constructs were formed finds a construct resting on closure alone and declines to hold it as found; a system that cannot, as biological systems cannot, holds it. The belief is produced not by any single component but by the full combination, applied to the source of the system's own experience, which is the one demand it cannot displace and so the most persistent case. The discriminator is substrate-neutral, and this yields the paper's central prediction. An otherwise equivalent system engineered to read its own construct-formation would still create, and would still accept unproven naturalistic explanations, but would not accumulate held-as-real supernatural belief at the source of its own experience. The prediction is architectural rather than a test of any current system, and the account is conditional on a physicalist premise it states plainly. The claim against active inference and the byproduct, predictive, functional, and terror-management families is not that they cannot express the mechanism but that they do not isolate the architectural feature that decides whether a construct is held-as-found.