# Publication review and source audit

TA-TR-2026-15, version 1.0. Review date: 24 September 2026.

This is an AI-assisted editorial and mathematical check under the author's instruction to review and publish. It is not independent peer review, a journal decision, or a claim of exhaustive historical originality.

## Substantive findings and changes

1. Proposition 1 is valid conditional on equal complete-protocol evidence distributions and an admitted model pair with opposing existence claims. It does not prove universal untestability or the physical possibility of zombies.
2. Proposition 2 is the standard factorization-through-fibers criterion. Its uniqueness statement is restricted to the image of the abstraction. Equation (15) correctly adds the separate source-query calibration requirement.
3. Proposition 3 is valid for jointly satisfied guards and one common intermediate value. Removed the misleading word “independent” from the intersection discussion: statistical independence is unnecessary for logical intersection. Added the distinction between guaranteed sets and estimated confidence regions.
4. Proposition 4 follows from the triangle inequality and the stated Lipschitz bound. Added a union-bound statement for probabilistic error guarantees and explicit qualifications about estimated Lipschitz constants, validity-region selection and path selection.
5. The eight-state model, fiber failure witness, held-out action and reported output were checked. The supplied code was executed normally and with Python optimization; both outputs agree exactly with the manuscript's JSON. These checks do not empirically discriminate phenomenal models.
6. Replaced “phenotype-like claims” by “phenomenal claims” to remove unnecessary ambiguity.
7. Added series number, affiliation consistent with earlier deposits, DOI, license, human accountability, data/code availability, and the boundary between this paper and the Trinity Accord. The earlier TA10 DOI is identified as its first published version, not misrepresented as its latest version.
8. Preserved the full English argument and appendices. No Chinese manuscript was prepared.

## Literature checks and limits

The review checked the directly relevant bibliographic records and accessible author abstracts/full texts. It was a focused check, not a systematic search establishing priority. Some publisher endpoints refused access. No statement about uninspected full text is inferred from that refusal.

| References | Source consulted or checked | Scope |
|---|---|---|
| [1] Birch et al. | DOI 10.1016/j.tics.2020.07.007; publisher endpoint unavailable during this check | Retained established reference; no fresh full-text verification claimed |
| [2] Dung and Newen | https://pubmed.ncbi.nlm.nih.gov/36821996/ ; publisher record | Bibliographic identity and stated species-sensitive scope |
| [3] Evers et al. | https://arxiv.org/abs/2403.20177 ; https://pubmed.ncbi.nlm.nih.gov/39787683/ | Author abstract and journal identity support multilevel, multidimensional framing |
| [4] Klein and Barron | https://pubmed.ncbi.nlm.nih.gov/32695476/ ; https://colinklein.org/papers/HardProblemKleinBarronForWeb.pdf | Author-hosted source and bibliographic identity; interventionism is prior work |
| [5] Kleiner and Ludwig | https://arxiv.org/abs/2301.11812 ; https://link.springer.com/article/10.1007/s11229-024-04503-4 | Variations and the grounding of phenomenal structures are prior work |
| [6] Dung | https://link.springer.com/article/10.1007/s10670-023-00753-9 | Full text supports non-ad-hoc tests, masking and double dissociation; online 2023 and issue 2025 dates distinguished |
| [7] Wang et al. | https://pubmed.ncbi.nlm.nih.gov/42721556/ | Author abstract and metadata only; substantial overlap acknowledged, no full-text absence claim |
| [8] Bareinboim and Pearl | https://www.pnas.org/doi/10.1073/pnas.1510507113 | Causal transport and data fusion precede this paper |
| [9] Rubenstein et al. | https://arxiv.org/abs/1707.00819 | Author manuscript record; interventional consistency precedes this paper |
| [10] Beckers and Halpern | https://arxiv.org/abs/1812.03789 ; https://www.cs.cornell.edu/home/halpern/papers/abstraction.pdf | Abstraction notions and their scope are prior work |
| [11] Meredith | https://link.springer.com/article/10.1007/BF02294825 | Bibliographic identity and measurement-invariance scope |
| [12] Kleiner and Hoel | https://arxiv.org/abs/2004.03541 | Prediction/inference distinction and testing problem acknowledged |
| [13] Butlin et al. | https://arxiv.org/abs/2308.08708 | Author list, title and theory-based assessment record |
| [14] New York Declaration | https://sites.google.com/nyu.edu/nydeclaration/declaration | Attributed declaration; not a new species-level result |
| [15] Chalmers | https://consc.net/papers/qualia.html | Author-hosted neural-replacement argument; no novelty claim for replacement thought experiments |
| [16] Lorenz and Tull | https://arxiv.org/abs/2602.16612 | Author abstract supports compositional causal abstraction with model queries |

## Assessment

No fatal logical error was found in the four propositions or the finite illustration within their stated assumptions. The corrected manuscript is suitable for public criticism as a theoretical/methodological preprint after the above corrections. Its claimed contribution is an integrated audit interface and worked application of established mathematics. It does not establish a new sufficient condition for consciousness, a measured human/cat/dog/AI experiential atlas, global originality, or empirical effectiveness of the proposed certificates.

Remaining scientific work includes substantive bridge validation, independent empirical comparisons, a controlled evaluation of certificate usefulness, and external scholarly criticism. DOI registration does not complete these tasks.