Title: Revisiting the double-binary-pulsar probe of non-dynamical Chern-Simons gravity

Abstract: One of the popular modifications to the theory of general relativity is non-dynamical Chern-Simons (CS) gravity, in which the metric is coupled to an externally prescribed scalar field. Setting accurate constraints to the parameters of the theory is important owing to their implications for the scalar field and/or the underlying fundamental theory. The current best constraints rely on measurements of the periastron precession rate in the double-binary-pulsar system and place a very tight bound on the characteristic CS lengthscale k_cs^{-1} <~ 3*10^{-9} km. This paper considers several effects that were not accounted for when deriving this bound and lead to a substantial suppression of the predicted rate of periastron precession. It is shown, in particular, that the point mass approximation for extended test bodies does not apply in this case. The constraint to the characteristic CS lengthscale is revised to k_cs^{-1} <~ 0.4 km, eight orders of magnitude weaker than what was previously found.