Why the Universe Looks Random (And Isn't): A Topological Phase Transition at the Edge of Observability

Einstein's field equations determine the curvature of spacetime but say nothing about its topology. The universe could be infinite or it could wrap around on itself — and general relativity cannot tell the difference. We introduce the Davis capacity C = τ/K as the first continuous, quantitative measure of whether a compact spatial topology is observable from within.

The key insight is gauge-theoretic: on a flat compact manifold, the Levi-Civita connection has zero curvature everywhere. This is why every statistical test of the CMB returns "random" — the connection is flat, so local probes see nothing. But flat connections on manifolds with non-trivial topology carry monodromy: accumulated phase around non-contractible loops, detectable only by winding around the entire manifold. This is the Aharonov-Bohm effect applied to cosmology.

Four theorems are proved:

• Theorem A (Flatness): The spatial connection is exactly flat (F = 0), explaining why all local CMB tests return null.
• Theorem B (Trichotomy): The parameter Γ = C/π classifies each topological cycle as determined (Γ > 1, visible), critical (Γ = 1, phase transition), or underdetermined (Γ < 1, invisible), with a computable topological horizon L^top = 4R_H / sin(|Φ|/2) — a new concept distinct from the particle and event horizons.
• Theorem C (Decomposition): The CMB two-point function splits via the Double Cover Principle S + d² = 1 into matched-circles signal (weighted by S) and large-angle power suppression (weighted by d²). The same topology that explains the missing quadrupole power also explains why matched circle searches find nothing — they are the S and d² faces of a single identity.
• Theorem D (Expansion in Fidelity): In ΛCDM, Γ(t) is strictly monotonically increasing. The universe does not expand into new space. It expands in fidelity: the same topology, seen more clearly. An ultimate topological horizon of ≈ 17.5 Hubble lengths is derived, beyond which topology is permanently undetectable — a cosmological no-go theorem analogous to the Bekenstein bound.

The computation for the cubic 3-torus at maximal holonomy yields a phase transition at L ≈ 2.0 Hubble lengths — coinciding with the L = 2–3 L_H range independently favored by Betti functional analysis of Planck satellite data (Aurich & Lustig, 2024). The critical scale factor is a* ≈ 1.08: the present epoch sits at the threshold of topological observability.

New observational predictions include:

1. Anti-correlated circle pairs in the CMB (phase offset Φ = π), which no published search has tested.
2. Phase-scanning matched circles searches that sweep all Φ ∈ [0, 2π) rather than only Φ = 0.
3. Permanent undetectability of topologies with all cycle lengths exceeding 17.5 Hubble lengths.

The framework connects to Weyl's equidistribution theorem (1916), Veech's two-circle problem (1969), and the Aharonov-Bohm effect (1959), identifying what each captures and what each misses. A companion paper applies the identical mathematical protocol to the decimal digits of π, demonstrating the universality of the Davis Method across gauge theory, cosmology, and number theory.

Keywords

cosmic topology, Davis capacity, flat connection, monodromy, Aharonov-Bohm effect, Double Cover Principle, CMB anomalies, topological horizon, expansion in fidelity, matched circles, 3-torus, ΛCDM, Planck

Part of

The Davis Field Equations research program (C = τ/K). Branch XI.