QUESTION-283: Can composing systems create an effective structure whose adequate specification needs a richer relational vocabulary than the parts needed separately (relations of higher arity, history, or a new kind) and, if so, under what conditions, and is the new vocabulary forced by the composite or only by the restricted class of descriptions in use?

Why it is a question

The owner’s review of the pull request, 2026-10-27, on THEORY-1090: “The cellular-sheaf example demonstrates that composition followed by reduction can force a change in the type of relational structure needed for an adequate description. That’s a remarkably direct mathematical analogue of the sort of event we keep encountering in Luria: the existing relation scheme cannot express what the composed organization requires.” And: “Can composition create an effective structure whose adequate specification requires a richer relational vocabulary than the parts required individually? I think that deserves to become a major research question in Nucleation.”

What the record holds:

• A demonstrated case. A star with stalk ℝ² at the centre and ℝ at three leaves, with maps first coordinate, second coordinate and their sum. Eliminating the centre leaves the boundary constraint v₁ = v₂ + v₃. No sheaf on the boundary vertices, all of whose conditions are pairwise, has exactly that space of consistent assignments (Hansen and Ghrist, §4.2, Fig. 4.1; THEORY-1090; CLAIM-480).
• A sufficient condition against it. With vertex stalks of dimension at most one, reduction stays within pairwise sheaves (Thm 4.3). For graph Laplacians the Schur complement is again a graph Laplacian (Thm 4.2).
• The history case. Eliminating a slow interior turns an instantaneous coupling into a relation over histories, a memory kernel (CLAIM-480). That is a richer vocabulary of another kind: relations among trajectories where the parts needed only relations among states.
• The record’s own case. The record’s amendments have added relation types and fields when a composed use of the grammar needed what it could not say. Examples are ADR-044’s corroborates beside rests_on and ADR-043’s warrant (CLAIM-419, CLAIM-479). CASE-178 measured how often. Whether those are instances of this question’s phenomenon, or only resemble it, is part of the question.

What it separates

1. Forced by the composite or by the class. The three-way relation is inexpressible within sheaves on the boundary graph. It is expressible in a hypergraph, or by keeping the interior vertex. A change of vocabulary may be forced by the composite’s behaviour, or only by having restricted the descriptions to a class that excludes the interior. The answer may differ between the mathematics and the record.
2. Effective or fundamental. The owner’s review set the limit: the result “does not by itself establish that a new fundamental physical interaction has arisen.” The three-way relation is the trace, through a hidden interior, of relations that are all pairwise.
3. Arity, history and kind. Higher arity (the sheaf case), history (the memory case) and a new kind of relation (the record’s amendments) may be one phenomenon or three.

What would count as an answer

• A characterization, for some class of compositional systems beyond scalar networks, of when reduction stays within the parts’ relational vocabulary and when it must leave it. Theorem 4.3 is one sufficient condition. A necessary one would be the first step.
• A physical or simulated system in which a boundary observer, confined to pairwise models, measurably fails to predict a composite until it admits a higher-arity or history-dependent relation, with the failure located as in CEGAR (NOTE-1231).
• For the record, a test of whether its scheme amendments followed compositions that its existing relation types could not express, or arose otherwise.