Concept study / Tests & research methodsConceptual illustration. Shapes, colors, and motion are not measured data.
The project needs several hypotheses, not one statement that must be accepted or rejected as a whole. The live register uses H01 for geometry’s proposed priority, H02 for recurrence across scales, H03 for retained information, and H04 for universal adaptation. H05 introduces the Alpha 1 parent system; H06 adds its proposed center and spiral; H07 concerns geometric routes; H08 concerns conscious experience. These labels organize distinct claims rather than established physical categories. The register must record which claims have quantitative models, which remain conceptual, and which currently have no accessible test. Evidence for one entry does not automatically support the others.
One claim per record
A record contains its precise sentence, scope, mathematical form, adjustable parameters, evidence links, competing explanations and revision history. “Geometry matters” is too broad to serve as a new prediction; “this boundary perturbation shifts this mode by this amount” can be tested. “The universe learns” requires a defined state, training history, readout and comparison process. If those elements are absent, the record remains conceptual. A paper demonstrating material memory belongs under a relevant mechanism record and must not be tagged as a direct observation of cosmic memory.
Connect dependencies explicitly
Some claims depend on others. Detecting local memory would support the feasibility of memory mechanisms in that material, but it would not establish that energy itself is the storage substrate. An Alpha 1 influence needs a coupling model before an astronomical residual can test it. Represent these links as requirements rather than persuasion: which equation must be true for this prediction to follow? If a dependency fails, mark downstream conclusions as unsupported. Preserve the earlier wording so a reader can identify whether a revision repaired a model or changed the question.
Record outcomes without rewriting history
Use outcomes such as not yet tested, inconclusive at stated precision, inconsistent in the tested domain, or provisionally supported for the specified observable. Do not replace an unsuccessful numerical prediction with a looser verbal claim and call the result a success. A revised hypothesis receives a new version, its own rationale, and fresh data where possible. The register should also distinguish a retrieved citation from a checked derivation and a reproduced calculation. These are different levels of review, each valuable when described accurately and attached to an identifiable artifact.
The mathematical lens
A proposed record structure for hypotheses; it is not a new law of nature.
Turn it into a question.
- Question
- Can two reviewers identify the same prediction and failure condition?
- Method
- Give independent reviewers the same record and ask them to extract observable, domain and decision rule.
- What would count against it?
- Disagreement about what would count against the claim means the record needs clarification.
A theory becomes easier to improve when its claims can fail separately.
Follow the original work.
These sources support the mechanisms and limits discussed here. They do not endorse Hill’s framework. Notes distinguish research proposals, reviews, and observations.
Continue the investigation
Edition 02 · 18 October 2026 · AI-assisted synthesis. See editorial method.