Concept study / Tests & research methodsConceptual illustration. Shapes, colors, and motion are not measured data.
A mathematical fractal can continue through arbitrarily many levels; a measurement cannot. Images have pixels, instruments have bandwidth, catalogues have selection limits and experiments end. These boundaries can flatten, bend or truncate a scaling curve. They can also hide a transition that contradicts an apparently universal law. Hill’s nested-system hypothesis makes the accessible range especially important: evidence within one finite range cannot demonstrate indefinite repetition. The proposed response is a deliberate finite-size analysis that varies sample size and observation window, then asks whether a physical parameter stabilizes or merely follows the limits of the dataset.
Vary the window deliberately
Subsample point counts, shorten time records and crop fields using a prespecified family of windows. Repeat the full analysis rather than reusing a fit selected from the largest dataset. A robust estimate should show understandable convergence or a modelled finite-size dependence. If its value tracks the window boundary, report that dependency. For correlated data, random deletion and contiguous cropping test different properties; use the version relevant to the measurement. Keep the full dataset available so the window study can be repeated without guessing which observations were removed.
Separate physical and instrumental cutoffs
A break in scaling may arise from a real characteristic length, or from blur, a detection threshold or a limited field of view. Compare instruments or processing resolutions when possible. In simulation, pass a known structure through an observation model and see which breaks appear. For a resonator, finite bandwidth can hide modes; for a memory trace, a short experiment can confuse slow decay with indefinite retention. A plateau extending to the last measured point supplies a lower bound on persistence under those conditions, not evidence that persistence is unlimited.
Make extrapolation conditional
If a model extrapolates beyond measured scales, show the extrapolation separately and state the assumptions that allow it. A proposed hierarchy from particles to Alpha 1 spans regimes where governing physics, observational access and meaning of distance differ. A finite empirical slope cannot carry that entire inference. The next useful contribution is a quantitative account of where the relation should fail or change, and a dataset that reaches that transition. Negative results at a predicted crossover are especially valuable because they discriminate between models that fit the same narrower range.
The mathematical lens
A schematic finite-size model. The form of correction must come from a justified model or be clearly labeled empirical.
Turn it into a question.
- Question
- Does the conclusion persist as the observation window changes?
- Method
- Repeat the pipeline across nested windows and realistic simulated observation limits.
- What would count against it?
- A claimed invariant that follows window size or resolution requires a revised interpretation.
Every scaling claim needs an explicit measured range.
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.