Einstein's Other Gravity and the Acceleration of the Universe
Eric V. Linder
TL;DR
The paper investigates curvature-free teleparallel gravity by promoting the torsion scalar $T$ to a general function $f(T)$, yielding second-order field equations and modified Friedmann dynamics $T=-6H^2$. It demonstrates how an effective dark-energy sector arises from $f(T)$ and presents two concrete models—power-law and exponential—that can drive late-time acceleration while reproducing GR at early times; a de Sitter fate is achievable in both cases, with parameter constraints ensuring observational viability. The work reveals connections to $f(R)$ gravity, higher-dimensional models like DGP, and scalar-tensor theories, and emphasizes the framework's ability to explain cosmic acceleration without curvature-based pathologies. Overall, $f(T)$ gravity offers a flexible, curvature-free route to modified gravity that preserves second-order dynamics and links to several broader gravity theories, making it a compelling alternative to explain the universe's acceleration.
Abstract
Spacetime curvature plays the primary role in general relativity but Einstein later considered a theory where torsion was the central quantity. Just as the Einstein-Hilbert action in the Ricci curvature scalar R can be generalized to f(R) gravity, we consider extensions of teleparallel, or torsion scalar T, gravity to f(T) theories. The field equations are naturally second order, avoiding pathologies, and can give rise to cosmic acceleration with unique features.
