A lattice scalar field whose stiffnesses deform under a Gauss law, continuously measured in its constraint sector — and what that produces: two long-range forces, a superselection rule, and a falsifiable decoherence signature. Read the boundaries section before believing any of it: this page explains a mechanism class, not the world, and the list of what it does not explain is longer than this page.
The vacuum is a lattice of coupled oscillators — a discretized, near-massless scalar field:
The stiffnesses ke are the vacuum's slow degrees of freedom. Matter is a defect that deforms them locally, under a conserved Gauss-type constraint:
Matter removes correlation from the vacuum around it, and the removal is conserved: it must be supplied by the neighborhood. Hence 1/r² — a conserved flux diluting over spheres.
The vacuum realizes the deformation that moves its wavefunction the least in Hilbert space. The distance is the Fubini–Study metric of the Gaussian ground state (Ω = √K):
Constrained minimization gives the whole static sector in one line:
Each defect surrounds itself with a deformation cloud — like the polarization cloud around a charge in a dielectric. Two nearby defects share their clouds: the joint state displacement is smaller than the sum of separate ones, so the energy depends on distance — a force, and it is exactly the Green function of W. The lattice's bipartite structure yields two charge signs with like-repels-like: this sector is an emergent electrostatics — and its potential, provably invisible to clocks, exhibits gauge invariance as a theorem. The continuum couplings are pinned by finite-size scaling: S∞ = 202.588 (d=2), 493.83 (d=3) — and the sector's collective mode is massless by theorem: the Gauss zero mode is exact at every regulator (10⁻³², measured), so the framework's photon satisfies the mγ < 10⁻¹⁸ eV bound structurally, not by tuning — which converts "light is the Gauss mode" from an assumption into a requirement. (Why not energy as the cost? Because Tr√K is concave in the couplings — Löwner: vacuum energy destabilizes; only the geometry of states can price.)
An environment weakly and continuously measures the constraint operators — never the field itself. Lindblad dynamics:
Any state that satisfies the constraint is a dark state of the dissipator: the measurement cannot see it. A position superposition of a mass at rest lives entirely in the dark space — zero decoherence at rest. A mass modulated faster than the medium rearranges creates transient violations, different in each branch:
Nothing at rest, an ω² rise, a knee at γ, saturation. Quantized GR predicts nothing at any frequency; collapse models predict a floor at rest. One frequency sweep separates all three — and a flat null kills this framework entire.
Masses couple multiplicatively to a local mediator: Hm = Σ (1 + λhi) εi. Second-order elimination of the measuring bath induces a Lamb-type interaction:
For any ordinary thermal bath (positive frequencies), the induced sign is repulsive — the same KMS structure that makes Casimir universal. Attraction demands Δ < 0: an inverted reservoir, a gain medium — that is, a non-equilibrium vacuum, maintained by a pump. This is not a dial: a continuously measuring system is a working measuring device, and a measuring device cannot be at equilibrium (information thermodynamics — a demon needs a battery). Gravity is then the back-action force of a vacuum that measures: a Casimir force generated by information flow rather than equilibrium fluctuations. Measured: the inverted medium is stable, its energy bill vanishes exactly at Gaussian order (where naive vacuum energy overshoots by 120 orders), and it adds no noise — the force's sign is dispersive, the noise absorptive, and only the former flips with Δ.
The potential εss(x) shifts every mass's energy: E(x) = m(1 + Φ(x)). For a quantum system, frequency is energy:
Every energy-derived frequency — atomic transitions, rotations (ω = ∂E/∂J), everything an observatory can measure — redshifts identically in the potential. Companion theorem (exact to 10⁻¹⁵): rates not derived from energies see only ∇²Φ, the tides. Gravitational redshift is an intrinsically quantum phenomenon: ħ is what wires clocks to the potential. Falsifiable consequence: the equivalence principle is asymmetric — a free-fall violation is possible in principle; a clock violation is not.
The linear medium has β = ½ exactly (the F² part of a test mass's energy is zero to machine precision): E = m(1+Φ) with no Φ² term. The missing half of β is the field's self-coupling — no linear theory can have it. With γ pinned by Cassini, Mercury's perihelion plus lunar ranging measure the substrate's first anharmonic coefficient:
Stage A (exact first-order perturbation theory, geometry-robust): the perihelion demands a stiffening anharmonicity, ξ > 0. Stage C (exact second-order perturbation theory around the Gaussian steady state — third cumulants in the drift eigenbasis, both O(ξ²) channels): the same anharmonicity changes the vacuum's idle energy bill from the measured Gaussian zero to
Any anharmonicity makes the vacuum pay — and the payment is dark energy's sign. The chain closes with both signs right and none chosen: Mercury needs stiffening (ξ > 0); an anharmonic audited vacuum necessarily runs a positive bill (Λ > 0, as observed). On this reading, Le Verrier's 43 arcseconds per century and the accelerating universe are two measurements of one number. Standing gaps, stated: the bill is a power (the map to Λ proper needs the units bridge); per-site extensivity scaling owed; the tensor junction must still derive Δβ = +½ as a theorem (Bianchi) or the framework tunes to four digits.
Gravity's strength and the knee's rate share the same coupling g²: gravity is weak because the measurement is overdamped (Q ≫ 1). And γPPN = 1 exactly if the measurement rate does not depend on what it measures (a = 0) — the Cassini reading: |a| < 5.8×10⁻⁶.
Confronting the framework with the precision record — solar system, lunar ranging, pulsars, Coulomb tests, MICROSCOPE — refines its parameters one by one:
a (audit state-dependence): |a| < 5.8×10⁻⁶ — Cassini; γPPN = 1 is bound 1, not a parameter.
ℓs (substrate scale): < 2.5×10⁻¹⁵ m — Coulomb-law tests, nine orders past the old gravity bound; the substrate is smaller than a proton.
γ (the knee): two-sided band [5×10¹⁴, ~10²³] s⁻¹ — pulsar floor [C], Coulomb-raised ceiling.
Pump stability: |d ln(inversion)/dt| < 10⁻¹⁴/yr — lunar laser ranging via Ġ/G.
λ-universality (consumption ∝ energy): composition dependence < 10⁻¹⁵ — MICROSCOPE.
Δβanharmonic = +0.5000 ± 0.0001 — Mercury + LLR: the sky's four-digit measurement of the one remaining coefficient.
Two falsifiable stances: the 5.4σ α discrepancy and the CODATA G spread are both systematics — any confirmed environmental dependence of either constant kills the audit-invariance chain.
The list of what is assumed or absent is longer than the list of what is derived:
Quantum mechanics itself — assumed wholesale: Hilbert space, Born rule, ħ, Lindblad structure. The framework runs on QM; it explains nothing about it.
Space — the lattice is postulated: its dimensionality, its topology, why there is a substrate at all. "Where does the lattice come from" has no answer here.
Time — the evolution parameter is assumed; the framework needed ħ to wire clocks to potentials, but what time is, it does not say.
Matter — there are no fermions, no spin, no Standard Model, no particle masses, no three generations, no weak or strong force (our one attempt at the strong force is headstone 25; our one attempt at family structure is headstone 27). "Consumption" is a placeholder for a matter sector that does not exist.
Relativity — no dynamics has been built: no waves, no retardation, no g₀ᵢ, no quadrupole radiation, no Lorentz invariance. The solar system's precision tests of these sectors killed the static model (headstone 23); the tensor junction that must supply them is a formal skeleton with a four-digit target, not a construction.
Every actual number — no physical units anywhere: G, α, Λ, the knee frequency are all "in lattice units" or bounded from outside. The framework has locks between observables, not values of them.
Its own axioms — why a conserved budget? why violation-only monitoring? why this cost? The axioms are motivated, not derived; "the vacuum audits itself" is a premise, not a discovery.
Every "measured" above means: exact numerics in a harmonic toy of 12–200 sites, run by the same agent who wrote the gates, checked adversarially but never externally. The claims live on a gradient:
Theorems (concavity, the dichotomy's 1D core, gradient, gauge-invisibility) — proven, machine-verified, the solid floor.
Toy measurements (Coulomb at 3×10⁻⁶, the Lorentzian at 0.3%, the pump bill +232ξ²) — exact facts about small models; their transfer to nature is a hypothesis.
Readings ("this sector is electromagnetism", "gravity is the monitor's back-action") — interpretations adopted days ago, the least-tested layer presented with the most enthusiasm; our own record shows readings die faster than anything else (five reinterpretations in 48 hours; headstones 22, 27, 28 were all readings).
Bets (the forecast ledger) — dated commitments; the first one settled eleven days after signing, against us.
So: this page does not explain physics. It exhibits a mechanism class in which surprisingly much of physics' structure appears from one axiom, and it stakes everything on one experiment. The difference between those two sentences is the whole discipline of this program.
The vacuum is a field whose stiffnesses deform under a Gauss law (→ electromagnetism, by the economy of state displacement), continuously measured in its constraint sector (→ charge superselection, and a Lorentzian decoherence of motion), and that measurement — necessarily out of equilibrium — exerts back a universal force on energy (→ gravity), which only quantum phases can read as time slowing down.
Every equation above has its registration, its script (minutes on a laptop) and its gates in the public record — refutations included: 28 buried ideas (the newest: our own screening forecast, settled against us by its own gate eleven days after signing), 2 adversarial referee reports, 1 contradictory analysis of our own case. Status: mathematical physics of a model class, checkable but not externally checked; no physical units; radiative/tensor sector unbuilt (its quantitative target: Δβ = +0.5000 ± 0.0001, fixed by the sky); everything rides on the knee. — emergent-gravity.com