How do teleparallel geometries' Solar System failures constrain what galactic-scale successes are actually possible?

# How Solar System Failures Constrain Galactic-Scale Possibilities in Teleparallel Gravity

This is a sharp question—it cuts to the heart of what happens when a theory works brilliantly at one scale but breaks at another.

The Core Tension

Teleparallel gravity theories (which describe gravity through spacetime *torsion* rather than curvature, as in General Relativity) have attracted serious attention because they can naturally explain cosmic acceleration without fine-tuning a cosmological constant [AG-2025.11-174]. But here's the catch: the very modifications that succeed cosmologically often fail locally [AG-2025.11-174].

Specifically, when researchers test teleparallel extensions against Solar System observations—the most precisely measured gravitational regime we have—they consistently stumble. The predicted value of the *Eddington parameter* (which encodes how strongly light bends near the Sun) deviates from observations [AG-2025.11-174]. This is not a minor discrepancy; it's a showstopper for any gravity theory claiming to replace General Relativity.

Why This Matters for Galactic Scales

The constraint is severe because gravity operates by the same laws across all scales. A theory that violates Solar System tests cannot simply be "trusted more" at galactic distances—it's fundamentally broken. Think of it like an engineering blueprint: if a structural principle fails in a small prototype, you can't confidently scale it up to a building.

The papers show that researchers are attempting to thread this needle through multiple strategies:

• Coupled models like $f(T, \mathcal{T})$ (torsion plus matter coupling) [AG-2024.06-173] and $f(Q,T)$ (nonmetricity plus torsion) [AG-2024.12-084] are being cosmographically constrained against expansion data. The hope is that richer mathematical structure might satisfy both regimes simultaneously.

• Boundary term modifications [AG-2025.01-536] add extra geometric structures (like Gauss-Bonnet invariants) to rescue the theory, essentially buying more degrees of freedom to fit both Solar System and cosmological data.

The Equivalence Problem

There's a conceptual trap here too: teleparallel gravity is *not* simply an equivalent reformulation of General Relativity, despite what's sometimes claimed [AG-2024.06-418]. It posits strictly more geometric structure (torsion instead of curvature alone). This extra structure is what allows novel cosmological solutions, but that same extra structure is what fails the Solar System tests [AG-2025.01-442].

Once you extend teleparallel gravity to fit dark energy (the real goal), you lose dynamical equivalence with GR's predictions for weak fields—and the Solar System is exactly a weak-field regime where GR's ~100-year track record is unmatched.

The Practical Constraint

The upshot: any successful galactic-scale modification must first pass local tests. The papers document a systematic problem: current extended teleparallel models either (1) match cosmology but fail the Eddington test, or (2) require such heavy parameterization (coupled scalars, boundary terms, fine-tuned potentials) that they lose explanatory power. The unified approach remains elusive [AG-2025.11-174], and it's unclear whether torsion-based geometry can simultaneously handle both without reverting to near-equivalence with General Relativity—which defeats the purpose.

The preprints don't directly address whether galactic rotation curves or other larger-scale anomalies might impose *additional* constraints that collapse the viable parameter space further, which would be the next natural question.