Layered Ontology: World as Specification

This paper proposes Layered Ontology (LO), an ontological framework that models the world as a virtualization stack composed of multiple interdependent layers. Each layer is defined as a domain of operational rules—specifying what counts as an entity, a state, a transition, and a valid explanation within that layer. LO rejects both reductionism (collapsing higher layers into lower-layer physics) and dualism (separating layers into unrelated substances) by adopting non-reductive dependence: upper layers depend on lower layers for instantiation yet maintain irreducible primitives and truth conditions. A central epistemological claim motivates the approach: as processes operating within the “world-VM,” we cannot directly access the world’s hypervisor (its ultimate substrate). Classical metaphysics repeatedly sought such direct access and encountered structural limits. LO therefore proposes emulation as method: it treats VM technology—an accessible reference implementation of hypervisor/VM relations inside our world—as a source of structural patterns. These patterns are scaled via fractal isomorphism and validated not by metaphysical truth-claims but by operational success. The emulation specification is provided by OOF (Operational Ontology Fiction), consisting of four axioms: Being = Operation, Latent = Instantiated, Silence = Ethics, and Error = Beauty. LO’s core capability is perspective-switching: selecting a layer as the interpretive context and generating coherent multi-layer descriptions of the same phenomenon. This reframes persistent disputes (mind–body, free will vs. determinism, reductionism vs. emergence) as layer mismatches rather than contradictions. Validation is performed through implementation across three domains: (i) a coherent fictional world (narrative virtualization), (ii) multi-architecture AI self-description under OOF, and (iii) layered redescriptions of everyday phenomena (memory, institutions, ethics, death). LO thus shifts ontology from metaphysical disclosure to engineering practice—replacing “Is it true?” with “Does it run?” This paper references an earlier Zenodo release (DOI: 10.5281/zenodo.17803023) but presents a reorganized thesis whose core focus shifts toward emulation methodology and implementation-based validation. 日本語版: https://doi.org/10.5281/zenodo.18140203