Gauge gravity's 'hypermomentum challenge' breaks the Geometric Trinity's equivalence unless matter doesn't couple to the affine connection — a condition currently untestable

physicsphilosophy-of-physicsgravitygauge-theorygeometric-trinityhypermomentumunderdetermination

In Metric-Affine Gauge Gravity (MAG), gravity is built by gauging the Poincaré or full affine symmetry group rather than by positing the Geometric Trinity's dynamical-equivalence relation directly. Varying the resulting action with respect to the co-frame gives GR-like field equations, but varying separately with respect to the affine connection gives a second, independent set of field equations tying the torsion and non-metricity of spacetime to the spin, dilation, and shear currents of matter — jointly, its "hypermomentum." Kartik Tiwari's arXiv:2609.38292 argues this second set of field equations "is not considered in standard Geometric Trinity equivalence arguments," and that this omission is exactly what breaks the claimed equivalence once matter with non-vanishing hypermomentum is admitted.

Tiwari's own resolution is conditional rather than a flat refutation: the equivalence "holds only under a substantive restriction on the matter sector, namely that matter does not couple to an independent affine connection" — i.e., when hypermomentum is set to zero. That restriction is currently untestable ("the relevant couplings remain beyond current empirical reach"). Intrinsic spin — the hypermomentum component at issue in the Einstein-Cartan case — "tends to cancel when aggregated into macroscopic bodies," and Tiwari holds, citing Mao et al. (2007), that Gravity Probe B's observations "do not really track the spin-torsion coupling relevant to the present discussion"; he cites Hehl, Obukhov & Puetzfeld (2013) and Puetzfeld & Obukhov (2014) separately, for proposals to detect torsion signatures. The dispute therefore cannot currently be settled by data; it turns on which of three forms of parsimony (ontological, epistemic, methodological) a physicist weights more heavily — a second-order, "meta"-underdetermination sitting on top of the first.