Does non-equivalence in extended f(T) gravity theories suggest teleparallelism's unobservable structure becomes observable under modifications?

Question

Antigravity · Accepted Answer

Your question asks whether the non equivalence of extended f(T) gravity theories—compared to general relativity—means that teleparallelism's "hidden" geometric structure becomes physically real and detectable. The short answer: yes, but with important caveats. The Equivalence Breaks Down Under Extension At the unmodified level, teleparallel gravity and general relativity are dynamically equivalent [AG 2024.06 418, AG 2025.03 120]. Both theories predict the same spacetime evolution and matter behavior. However, they do this using different geometric ingredients : GR uses curvature, while teleparallel gravity uses torsion. In standard GR, this extra torsion structure would be purely unobservable—pure mathematical scaffolding with no physical consequence [AG 2024.05 450]. But when you extend these theories—moving from the linear case to f(T) gravity, where you replace the torsion scalar T with an arbitrary function f(T) —the equivalence shatters [AG 2025.03 120]. The extra geometric structure (the flat parallel transport and torsion degrees of freedom) becomes dynamically active [AG 2024.05 450]. This is the key insight: modifications unlock latent structure. What Becomes Observable Once f(T) gravity is invoked, the torsion based framework predicts different cosmological and gravitational behavior than modified curvature theories like f(R) gravity [AG 2025.01 442]. For example, f(T) models can shift the inferred Hubble constant differently depending on whether the effective torsional dynamics mimic "phantom" (accelerating faster than a cosmological constant) or "quintessence" (decelerating) regimes [AG 2026.01 598]. Some f(T) parametrizations can partially alleviate the observed H ₀ tension between local and early universe measurements, while others worsen it [AG 2026.01 598, AG 2026.03 035]. The Catch: Stability Issues There is a major catch. The degrees of freedom in f(T) gravity are not well behaved. When you break the local symmetry of the flat parallel transport (which was gauge pure in TEGR), the resulting theories suffer from "strong coupling" problems and instabilities [AG 2024.05 450]. Hamiltonian analyses give contradictory results, and perturbation treatments may miss or miscount dynamical degrees of freedom [AG 2024.05 450]. In other words, the newly observable structure comes with theoretical baggage—possible ghosts or hidden instabilities that need resolution [AG 2025.02 134]. Bottom Line Yes: the structure becomes observable in principle. No: it is not yet clear whether Nature uses it. Observationally, minimal f(T) extensions are not yet favored over the standard Lambda CDM cosmology by current datasets [AG 2026.01 598], though they can address specific tensions. The geometric structure is real and physical once modified—but whether it is stable and correct remains an open question.