| Mathematical formalisms can be transferred as modeling lenses without claiming the target domain is made of the source-domain stuff. | Wigner on mathematical usefulness, Jaynes on probability as logic, and Khrennikov’s quantum-like modeling line. | Treat QL as a math-lens transfer: name the useful structure, the ordinary FPF pattern, and the local stop before any claim requiring additional evidence or authority. |
| Quantum-like open-system and instrument formalisms can model state and probe interaction without physical quantum ontology. | Basieva, Khrennikov, and Ozawa and arXiv, plus Khrennikov on open systems. | Keep QL-NQ central and use QL only where probe, instrument, open-information-system update rule, probe frame, export admissibility, or state export cue changes the admissible reading. |
| Question order, contextual judgment, and instrument-like operations are practical cues, but not automatic proof that QL is necessary. | Quantum instruments for question-order effects, Quantum Cognition, and process-theory non-exclusivity. | Use QL-lite when order/frame/probe effects change the result; keep classical instrument, Bayesian, causal, and ordinary measurement rivals live. |
| Same-content-looking measurements under different probe or measurement frames should not be silently treated as the same random variable or as jointly distributed. | Contextuality-by-Default. | Use C.26 only when the exact frames change variable identity, joint availability, or admissible comparison and a named obstruction survives ordinary measurement and Bridge patterns; otherwise keep those ordinary subject patterns. |
| Viability and active sensing often mix reading and acting, but ordinary control and measurement patterns remain primary. | Free-energy and quantum-cognition link, physiological regulation and FEP, active inference behavior, and smart-building active inference. | For viability cases, name sensors, probes, actuators, and envelope variables first; retain QL only for remaining probe, frame, export, or coarsening cue. |
| Boundary and DDD-locality questions are already disciplined by ordinary architecture practice. | Computational boundary of a self, Markov blankets of life, Azure domain analysis, and DDD 2025 SLR. | Apply direct boundary, local-sense, work, model-use, interface, and Bridge subject patterns first. If Markov-blanket wording is present, recover its exact claim and subject pattern; retain C.26 only where a named probe, order, comparison, export, or state-reading obstruction remains load-bearing. |
| Low-bit, tokenized, compressed, geometric, or neural representations may be useful shortcuts without being QL activation. | 1-bit LLMs, implicit continuity in language models, emergent quantumness in neural networks, and covariant gradient descent. | Keep implementation substrate, geometry, compression, and representation shortcuts in ordinary FPF patterns unless a declared QL cue changes the admissible use. |
| Unknown alternatives and regime movement are search/generation problems, not QL claim authority. | Open-endedness and quality-diversity through AI feedback. | Use QL only to mark a suspect frame; apply search or regime patterns to generation of alternatives. |