Binary Pulsar Timing and the Non-Propagating Character of Einstein-Cartan Torsion

Binary pulsars test gravitational radiation reaction with per-mille and better precision, and they are often proposed as torsion probes. This paper works out what they actually constrain. In Einstein-Cartan theory torsion is algebraic — it has no wave solutions, carries no energy to infinity, and therefore contributes exactly zero to orbital period decay: the theory's prediction for the double pulsar is precisely the general-relativistic one, which the observations match. The pulsars' power falls on extensions in which torsion propagates as an axial mode: there, a difference between the two stars' torsion charge-to-mass ratios radiates dipolarly, enhanced over quadrupole radiation by the inverse square of the orbital velocity. From the double pulsar's agreement with general relativity at a fractional excess below 1e-04, with an orbital beta of 2.09e-03 and a stated charge-asymmetry factor of 0.1, the torsion charge-to-mass ratio in gravitational units is bounded at kappa below 6.60e-05; a Hulse-Taylor-class system at a fractional 2e-03 gives 2.95e-04. The paper emphasises the logical structure: the pulsars do not measure Einstein-Cartan torsion and cannot confirm it; they certify its non-propagation by continuing to agree with a theory in which all radiation is metric, and they fence every propagating-torsion model at the kappa levels computed. Computed from stated assumptions, not measured.