Necessary Conditions for Primary Interoceptive Sentience in Continuous Substrates: A Falsifiable Programme

This paper asks a narrower, more tractable question than "what is consciousness": which physical substrates, if any, could in principle support primary interoceptive sentience — the most basic capacity of a system to register its own physiological state as good or bad for it? Rather than treating this as a purely philosophical question, we formulate it as a set of necessary physical conditions for continuous neuromorphic substrates (e.g. diffusive and Mott memristor networks), each paired with a falsifiable, measurable test. The programme makes no stronger claim than that: it does not propose a theory of consciousness, and it does not claim that the conditions it identifies are sufficient for sentience — only that they appear necessary, argued a posteriori from the biological case. A substrate that fails any test is not a candidate; one that passes all of them remains a provisional candidate, not a confirmed one. This is the distilled entry point of the P0_Distilled research corpus — the shortest, most self-contained paper in the series. The full research programme (Paper I), its extension to minimal self-agency (Paper II), an external critical assessment (Paper III), and two companion papers (philosophical and technical) are cross-linked via the related-works metadata below. Version 2 (September 2026): corrected incomplete literature citations in the related-work comparison (real author names added where missing; two entries that could not be independently verified were removed, along with the paragraphs describing them); softened an overclaimed originality statement to explicitly note the limited scope of the underlying literature survey.Version 3 (September 2026): corrects a file mix-up in Version 2, which was mistakenly published containing content from an unrelated internal exploration (an experimental reframing using Laudan's philosophy-of-science framework, distinct from this paper's Lakatosian structure). This version restores the correct file; no other content changed relative to the intended Version 2.Version 4 (September 2026): adds, in compact form, the load-bearing results of the two companion papers — a summary of the three-stage argument for Postulate 2 together with the four weaknesses identified in the critical assessment (§6), and the Monte Carlo power-analysis results qualifying the N = 50 sample size for the finite-size scaling criterion (§5); adds the observation that only H7 bears evidentially on Postulate 2 (§1); adds the survey-limitation caveat to the Conclusions; corrects the survey size (five works, not seven), the abandonment-criteria count (five, not four) and two cross-references; adds Zenodo DOIs for all companion papers.Version 5 (September 2026): adds a related-work section on nonequilibrium brain dynamics as a marker of the level of consciousness (Sanz Perl et al. 2021; Gilson et al. 2023; G-Guzmán et al. 2023; Sekizawa et al. 2024; Kringelbach et al. 2024), distinguishing it from the programme's focus on valence; adds a caveat on Griffiths phases to the critical scale-invariance condition; adds a note on the thermodynamic consistency of estimating Ψ(t) with Langevin models in the nonlinear regime; clarifies that Postulate 2 is a restricted variant of dual-aspect monism whose arguments are closest to an a posteriori identity thesis; corrects the statement that the No-Go Theorem is a preprint (it is published in Neuroscience of Consciousness).Version 6 (September 2026): adds related work on substrate-dependent views of consciousness and on the indicator-based approach (Butlin et al. 2023; Seth 2025; Findlay et al. 2024; Wiese 2024); discusses the unfolding and substitution arguments against causal-structure criteria and why the operational demarcation is not their standard target (Doerig et al. 2019; Kleiner & Hoel 2021; O'Reilly-Shah et al. 2026); notes that the structural isomorphism behind Postulate 2 faces the Newman problem (Kleiner 2025); adds two results on nonequilibrium brain dynamics (Lynn et al. 2021; Stikvoort et al. 2025); and places the programme's outcome in the precautionary vocabulary of Birch (2024).Version 7 (September 2026): specifies the second and third demarcation conditions in a form that can be estimated: quantised recordings are dithered, and the Markov threshold is preregistered per substrate and bounded by the record length, since a fixed value of 500 is not estimable with k-nearest-neighbour estimators; the tolerance of the third condition is stated as a minimum effect size, complemented by a significance threshold on linear surrogates and a minimum record length set by a power analysis. It also adds a reference to the full statement of the programme and corrects the titles of the companion papers.Version 8 (September 2026): corrects the description of the technical companion. From Version 4, this paper summarised Monte Carlo power-analysis results qualifying the N = 50 sample size for the finite-size scaling criterion; those results were not the output of an executed simulation and are withdrawn (technical companion, Version 5). Until the designed study is run, the criterion rests on the heuristic estimate of the full programme, and N = 50 is a working assumption for it.