Can torsion-based or nonmetricity-based theories produce distinct lensing signatures beyond the curvature formulation?

# Torsion and Nonmetricity in Gravitational Lensing

Yes, torsion and nonmetricity can produce observationally distinct lensing signatures that depart from curvature-only general relativity—though the effects depend sharply on the scale and geometry you're probing.

The Topological Lensing Regime

The clearest deviation comes from pure torsion without curvature. In a $(2+1)$-dimensional spiral dislocation spacetime, torsion alone creates a "topological exclusion zone" around the defect core [AG-2025.07-398]. Rather than smooth bending, null rays experience a finite turning radius set by the torsion parameter, and the deflection angle *decreases* monotonically as torsion increases—the opposite of curvature-driven lensing. This happens because torsion modifies the affine structure (the way paths are defined in spacetime) rather than the metric distances [AG-2025.07-398].

Crucially, this torsion-mediated deflection is *purely topological*: it doesn't arise from spacetime curvature at all. The same geometry also acts as a frequency-filtering waveguide for scalar waves, with the refractive index diverging near the core due to torsion-induced geometric terms [AG-2025.07-398]. This suggests that observations of photon and wave propagation together could distinguish torsion from curvature.

Nonmetricity Effects

Nonmetricity—where the connection fails to preserve distances under parallel transport—also contributes new degrees of freedom. In the most general formulation with both torsion and nonmetricity as independent dynamical fields, a naturally light scalar-like distortion field can emerge, mixing with the Higgs sector [AG-2026.03-265]. This adds new particle physics to the lensing picture, though the optical signatures remain to be worked out observationally.

Strong-Field Measurements Beyond Local Tests

A global Gaussian bending measure, defined using differential geometry rather than any particular gravity theory, can characterize lensing and bending of photons, neutrinos, and cosmic rays in stationary spacetimes [AG-2024.08-064]. This framework potentially allows local strong-field experiments—such as probing black hole shadows and lensing near rotating black holes—to distinguish between torsion-based, nonmetricity-based, and curvature-only theories [AG-2024.08-064].

The Comparison with Standard Tests

For comparison, modified gravity theories that keep the standard metric-curvature framework (like $f(R)$ gravity and nDGP models) produce lensing signatures that are *quantitatively* different from general relativity—larger Einstein radii, higher optical depths, and stronger magnification—but *qualitatively* similar in structure [AG-2024.11-168]. They enhance lensing, especially at high redshift [AG-2024.11-168].

Torsion and nonmetricity theories, by contrast, can produce *qualitatively* new effects: topological deflection zones, refractive index profiles absent from curvature, and coupling to new scalar fields. These are in principle more distinctive, though they may require probing regimes (cosmological defects, quantum gravity scales) where curvature effects are subdominant.

The distinction hinges on whether you have access to strong-field regions where torsion or nonmetricity dynamically dominate, or whether these fields couple to observables like wave frequency or particle spin in ways curvature alone cannot match.