The persistent lack of convergence in consciousness research is approached here as a structural signal that the field lacks a principled way of identifying the appropriate level of explanation, rather than reflecting only a shortage of data. The central proposal is that major ontological positions in consciousness science may be understood as epistemic terminators, i.e., conceptual commitments that halt an otherwise unbounded regress of explanation by stipulating what is to be treated as primitive, thereby converting open explanatory targets into assumptions that can appear theoretically complete.This termination thesis is developed by integrating Uexküll's Umwelt theory, Kantian and neo-Kantian constraints on knowledge, and predictive-processing approaches that treat perception as model-based inference. On this view, commonly invoked ontological categories — such as matter, information, intrinsic cause–effect power, the physical, or the mental — may be products of organism-relative construction and may not provide an Archimedean standpoint outside the biological apparatus that generates them.The framework is not intended to be anti-realist, as it remains bounded by measurable biological constraints rather than unconstrained metaphysical identity claims. PET findings establish a restricted metabolic range associated with reportable consciousness, while perturbational complexity measures define a validated boundary separating conscious from unconscious regimes across sleep, anaesthesia, and disorders of consciousness. These observations motivate a constraint-based approach in which consciousness may be understood as a metabolically sustained regime of large-scale, history-dependent integration that remains open to empirical disconfirmation.Within this perspective, the apparent incompatibility of leading theories may arise because each places an epistemic terminator at a different level of biological organization and generalizes from that level as if it were exhaustive. The aim is therefore not to eliminate such commitments, but to render them explicit and empirically tractable, thereby supporting more cumulative and empirically tractable progress in consciousness science.
Shivashanmugam et al. (Tue,) studied this question.