Definitional anchors

Most signals in the Observatory earn their verdict by surviving the statistical battery: a deterministic set of tests against the candidate’s data, followed by the Devil’s Advocate pass and the Whewell rubric. A handful of signals cannot be evaluated that way, because there is nothing for the battery to falsify. These carry the DEFINITIONAL verdict.

A definitional anchor is a signal whose content is a foundational fact — a physical, mathematical, or biological definition, constant, or convention. The SI second is defined; it is not a hypothesis that could fail an out-of-sample hold-out test. General relativity, as the framework the corpus uses to reason about gravitation and time, is a settled theory, not a candidate finding. Turing’s reaction-diffusion model of morphogenesis is a foundational mathematical result. Running a Bonferroni-corrected significance test or a phase-randomised surrogate control against any of these would not be wrong so much as meaningless: the batteries assume a hypothesis that could, in principle, turn out to be noise. A definition cannot.

So the verdict is honest about its own basis. DEFINITIONAL means validation by established literature and convention — not battery passed. It is deliberately kept distinct from CONFIRMED, which implies a battery was run and cleared. A definitional anchor that were stamped CONFIRMED would misrepresent how it was checked. Treating the two as interchangeable fails the audit gate.

Why such signals exist in the corpus

Definitional anchors are not filler. They are the fixed reference points other signals are measured against. A frequency-biology signal that claims a mechanism operates at a particular rate needs a defined unit of time to make that claim precise; an orbital-forcing signal needs the Milankovitch cycles as a stable backbone. Carrying these foundations in the corpus — rather than leaving them implicit — means every signal that depends on one has a traceable link to it, and the dependency is visible to the audit gate rather than assumed.

Keeping them as their own verdict tier also protects the kill-rate statistics. A definitional anchor can never be killed and never needs to be confirmed by battery; folding it into the CONFIRMED count would quietly inflate the confirmation rate with signals that were never at risk of failing.

Example anchors

The corpus carries definitional anchors across physics, astronomy, biology, and engineering convention. A representative selection:

• The SI second — the base unit of time, defined by convention (BIPM) rather than measured.
• General relativity — the settled framework for gravitation and the relativity of time, used as a reasoning substrate across the corpus.
• Turing morphogenesis — the reaction-diffusion model of biological pattern formation, a foundational mathematical result.
• Milankovitch orbital cycles (~100 kyr) — the defined orbital-forcing periodicities that other climate signals are referenced against.
• Axial precession / the Great Year (~25,772 yr) — a defined astronomical periodicity.
• The Chandler wobble (~433 days) — a defined periodicity of Earth’s rotational pole.
• The human hearing range (~20 Hz–20 kHz) — a defined sensory bound used to scope acoustic signals.
• Trichromatic colour vision — the defined three-channel basis of human colour perception.
• Power-grid synchronisation (60 Hz) — an engineering convention, not an empirical finding.

Each of these is a fact the corpus stands on, not a claim the corpus is testing. That distinction is the whole point of the tier.

See the Methodology Taxonomy for where DEFINITIONAL sits in the full verdict set, and the verdict tiers page for the canonical dossier mapping.