Minimal • Falsifiable • Tabletop-discriminable

Variance-Sourced Quantum Gravity (VSSG): a testable proposal

This page is a single-source, website-first research note for Project Curvature. The central claim is operational: two quantum preparations can share the same ⟨ρ̂⟩ yet differ in Var(ρ̂), and gravity may—under this proposal—distinguish them.

Weak-field limit Superposition vs mixture discriminator Single new parameter ρ* Versionable research note

Summary

Standard semiclassical gravity sources the classical field from the expectation value ⟨T̂μν⟩. VSSG introduces a deterministic correction sourced by irreducible quantum fluctuations (variance). This yields a concrete discriminator: coherent superposition and incoherent mixture can share the same ⟨ρ̂⟩ but differ in Var(ρ̂), producing a measurable difference in gravitationally mediated phase (weak-field).

New ingredient Deterministic variance-sourced correction with a single parameter ρ*.
Key discriminator Same ⟨ρ̂⟩, different Var(ρ̂) ⇒ different Φ (Newtonian limit).
Test path Gravity-mediated phase / entanglement style experiments (tabletop).

Quick definitions

  • ⟨ρ̂⟩: expected mass density in a quantum state
  • Var(ρ̂): ⟨ρ̂²⟩ − ⟨ρ̂⟩² (irreducible uncertainty)
  • ρ*: new density scale controlling deviation strength
  • Φ: Newtonian gravitational potential (weak-field)
Goal: extract a clean, unavoidable prediction and enable experimental bounds on ρ*.

Core postulate (weak-field)

In the Newtonian (weak-field, slow-motion) regime, the gravitational potential satisfies:

∇² Φ(x) = 4πG [ ⟨ρ̂(x)⟩ + (1/ρ*) Var(ρ̂(x)) ] Var(ρ̂) := ⟨ρ̂²⟩ − ⟨ρ̂⟩²
Operational meaning: two preparations with identical ⟨ρ̂⟩ but different Var(ρ̂) produce slightly different Φ. The cleanest discriminator is coherent superposition vs incoherent mixture.

Practical note: for pointlike masses, define a smeared density operator ρ̂σ over a small length σ to keep Var(ρ̂σ) finite. The platform choice controls σ (effective localization scale).

New testable prediction

For a “two-lump” equal-weight superposition, the variance contribution scales like:

Var(ρ̂σ) ∝ (m² / σ³) · p(1−p) (maximum at p = 1/2) Leading fractional phase correction scaling: φV / φGR ~ m / (ρ* σ³)
  • Heavier mass m increases sensitivity
  • Tighter localization (smaller σ) increases sensitivity strongly (∝ 1/σ³)
  • Null results bound ρ* from below

What a bound looks like

If an experiment measures gravitational phase to fractional precision ε without detecting a deviation:

ρ* ≳ (m / σ³) · (1 / ε)

That turns a foundational idea into a real experimental constraint.

First platform target

Start with gravity-mediated phase / entanglement style experiments: two masses placed into spatial superpositions, where Newtonian gravity induces branch-dependent phases. The discriminator is: coherent superposition vs incoherent mixture at the same ⟨ρ̂⟩.

Why this platform first It directly tests whether gravity distinguishes coherence (superposition) from mixture — exactly what Var-sourcing changes.
Measurement Branch-dependent phase shifts; optionally an entanglement witness if implemented as a full mediator test.
Next deliverable A worked two-branch model with smearing scale σ, producing a numeric forecast for ρ* bounds.

Implementation details (m, separation, coherence time, σ definition) will be tailored once you pick a target apparatus (e.g., levitated optomechanics, matter-wave interferometry, or hybrid resonator systems).

Kid version (plain language)

Imagine gravity is like a “pull” that comes from mass. Usually we say gravity only cares about where the mass is on average. But quantum things can be in two places at once.

New idea: gravity might care about how “uncertain” the mass is—not just where it is on average. So a mass that is truly in two places at once could make a tiny bit different gravity than a mass that is just “sometimes left, sometimes right.”

We test it by putting tiny masses into a quantum “two-places” state and measuring whether gravity creates a slightly different phase effect than the normal average-gravity prediction.

Notes & next upgrades

  • Covariant embedding: extend the postulate from Newtonian density to a stress-energy formulation.
  • Regularization: define ρ̂σ precisely for a platform (σ = localization scale / mode width / wavepacket width).
  • One clean number: compute an explicit bound forecast on ρ* for a target apparatus.
  • Versioning: add a changelog and PDF snapshot for citations.
Responsible framing: this is a test-focused proposal. The “win” is a clean bound or detection, not a headline.
Project Curvature — Draft v0.1
Tip: create a /versions page and copy each update as v0.2, v0.3… so people can cite a stable snapshot.