When Does Information Become an Affective Stimulus? TM–BM Resonance, Functional Stimulus-Status, and the Internal Locus of Affective Operativity

Abstract

This preprint introduces the TM–BM Affective Stimulus-Locus framework, a computational account of when represented information becomes affectively operative for a particular observer. TM–BM formalizes the processing architecture as: X → T → R → Z → E → I → A, where external input (X) is represented (T), evaluated through observer-specific resonance or match (R), assigned a functional affective stimulus-status (Z), and subsequently expressed through affective meaning (E), implementation/readiness state (I), and measured activation or response (A). The central contribution is deliberately narrow. TM–BM does not claim novelty for observer-dependent relevance, learned stimulus function, contextual modulation, incentive salience, resonance, or gating. Instead, it proposes that affective operativity itself warrants a separately typed computational status, Z, distinct from representation, graded relevance or resonance, downstream affective content, and measured response. This commitment is falsifiable: Z must demonstrate predictive, distributional, mediational, or causal value beyond well-specified continuous and non-Z alternatives, or its independent explanatory status should be rejected. Four studies test progressively different implications of this proposal. Studies 1 and 2 examine computational identifiability and distributional distinguishability under explicit rival models. Study 3 tests pre-outcome physiological reassignment using real human data. Study 4 examines disruption and subsequent repair of an evaluative state following misleading information, with behavioral evidence treated as primary and EEG differentiation interpreted conservatively. The accompanying supplementary material provides extended mathematical specification, methods, robustness analyses, source provenance, identification boundaries, and falsification criteria. The reproducibility archive contains the analysis notebooks, derived results, provenance documentation, manifests, and cryptographic checksums. Third-party raw datasets are not redistributed.