Cross-Substrate Phenomenal Comparison: A Typed Transformation-Transport Framework under Existential Uncertainty

Hongju Liu · TA-TR-2026-15 · Version 1.0 · 24 September 2026

Published English-only preprint; substantially AI-assisted, human-responsible, not peer reviewed.

Abstract

Comparing human, non-human animal, and artificial experience requires distinguishing whether a target has phenomenal experience from which aspects of its experience would be comparable if it does. Multidimensional profiles, behavioral similarities, and mechanistic correspondences provide important evidence, but do not by themselves settle measurement semantics, information loss, or the scope of cross-system inference. This paper proposes a typed transformation-transport framework whose basic comparison unit is a system-in-regime undergoing a specified intervention. Claims are distinguished by their targets: phenomenal existence, character, valence, perspectival organization, continuation, and token multiplicity. No universal human-centered phenomenal coordinate system is required, although each comparison requires an explicit local relational interface. The principal methodological object is a transport certificate separating observable causal correspondence, conditional phenomenal interpretation, and evidential updating about existence. Four propositions establish conditional non-identification under observational equivalence, a fiber criterion for query-preserving abstraction, guarded composition of set-valued constraints, and error propagation for metric interfaces. An exhaustively checked eight-state example shows that exact interventional correspondence can preserve one candidate phenomenal query while concealing a distinction relevant to another. A held-out intervention reveals the resulting model disagreement, but cannot empirically adjudicate it without an appropriate measurement connection. Bounded research protocols and a twelve-case conceptual regression suite operationalize the framework for human–animal–AI comparisons. The contribution is an integrated method for specifying and auditing transport claims, not a new sufficient condition for consciousness. In particular, the framework neither treats non-entailing evidence as worthless nor mistakes a formally valid conditional for an empirically established fact about experience.

Scope and research boundary

The paper distinguishes existence, character, valence, perspective, continuation and multiplicity claims. Four elementary propositions define obligations for conditional comparison; the constructed eight-state example is reproducible. It does not measure the experiences of humans, cats, dogs or AI, and does not establish a sufficient condition for consciousness or the empirical effectiveness of transport certificates.

This is first-party Adjacent Research Program scholarship (L4). It does not define, amend, validate or authoritatively interpret the Trinity Accord’s three Bitcoin Originals, and is not independent corroboration of earlier papers. The author is the project’s Guardian; substantial ChatGPT assistance is disclosed. This new paper has its own DOI and preserves earlier DOI records.

CC BY 4.0 applies within the author’s rights. DOI registration and file-integrity checks are not peer review or search-indexing guarantees.

Research series