What it is
General relativity is treated as an effective theory valid at large distances, and adding terms quadratic in curvature, a common way to probe physics at very short distances, generically brings in a massive spin-2 'ghost', a mode that makes the theory non-unitary; enlarging the geometry with torsion gives a ghost-free version (as the authors' preprint explains). The authors compute the gravitational waves both versions emit and confront them with observations. In the ghostful theory, when the extra modes become light, the massless graviton and the extra massive spin-2 ghost and scalar modes cancel, so the quadrupole formula of general relativity is not recovered, in direct conflict with precision measurements of binary stars. Using pulsar binaries and the neutron-star merger GW170817, the authors derive a lower bound of 4.6 × 10⁻¹² eV on the mass of the extra modes in the ghostful theory, and mass-dependent constraints on the coupling constants of the ghost-free one.
Why it matters
Gravitational waves can now test extensions of Einstein's gravity, and the authors call this the first gravitational-wave test of fourth-order gravity. The ghostful theory already had a theoretical problem, its non-unitarity (as the preprint notes); the published result adds an observational one, since with light extra modes it cannot reproduce the binary measurements that general relativity matches. The ghost-free theory instead recovers general relativity as its two couplings go to zero, whatever the mass of the modes (the preprint's statement), so the data cap how large those couplings can be at each mass. The authors say the approach can be extended to alternative gravity theories.
Underlined numbers link to their source. Every metric and quoted figure is listed under Sources and data below.
Filed underPulsars and Gravitational Waves Research, Cosmology and Gravitation Theories, Noncommutative and Quantum Gravity Theories
