PULSE evidence map¶
This map answers one concrete question: what authority does each PULSE component have today, and what does it not authorize us to claim? It does not produce a total score or turn abundant documentation into validation.
| Component | Status | Admissible support | Decisive boundary |
|---|---|---|---|
| PULSE identity and definition | Constitutional | PULSE DNA and Identity | Owner authority, not empirical effectiveness |
| Decision First and PDAMR | Constitutional and practice | DNA, Identity and complete author POV | No external performance comparison exists |
| Decision Circle | Constitutional | DNA and Identity | The cycle organizes practice; it does not prove learning |
| Five verbs | Practice synthesis | Author POV and constitutional sources | They are not a universal sequence |
| Human-in-Control | Constitutional principle | DNA, Identity and BoK controls | Documenting it does not guarantee effective control |
| Decision Experience | Candidate contract | Derived architecture and specification | Requires evaluation with real users and tasks |
| L0–L7 architecture | Candidate derived architecture | SRC-ARCH-001 |
Not approved doctrine or a validated architecture |
| Five levels of analytics | Author pedagogical framework | SRC-JF-006 |
Not a universal maturity scale |
| Interoception, allostasis and decision organism | Metaphors and hypotheses | Derived architecture | Not causal mechanisms or biological equivalence |
| Longitudinal record | Learning hypothesis | DNA, architecture and Measurement Record | Recording does not prove behavior or outcome change |
| B2B case | Instrumented, unexecuted design | MVP design and simulated data | No real action, observed outcome or attribution |
Claim compatibility¶
- Constitutional sources may govern PULSE definitions.
- An author POV may explain lineage or purpose.
- A derived architecture may support a candidate design hypothesis.
- None of those classes can establish superiority or effectiveness by itself.
SRC-JF-001andSRC-JF-003remain blocked because their copies are truncated.- An AI-assisted synthesis does not replace an original source or validate the architecture.
Open tensions¶
Practical framework versus technology ladder¶
The five-level framework may help teach analytics capabilities, but it conflicts with Decision First if interpreted as a requirement to “move up” before defining the decision. It remains a pedagogical taxonomy, not a mandatory maturity path.
Biological metaphor versus causal explanation¶
Interoception and allostasis can generate useful questions about signals, feedback and adaptation. The risk is turning verbal similarity into a mechanism. Any derived prediction must be observable without requiring the metaphor to be literally true.
Integration versus originality¶
PULSE may create value through selection and coordination even when many components have antecedents. Assembly originality, theoretical novelty and effectiveness are separate claims requiring separate evidence.
Next evidence threshold¶
Before an effectiveness claim is strengthened, a pilot needs:
- a decision, population and comparator defined in advance;
- a verifiable baseline and evidence quality;
- operational human authority and stop conditions;
- outcome, process, adoption and risk measured separately;
- enough follow-up to observe action and learning;
- publication of adverse, null and cost results.
See the machine-readable
governance/registries/evidence-profiles.yml for complete provenance, rights
and profiles.