Comment on “analysis of data on polarized lepton-nucleon scattering”

We comment on the suggestion by Ellis and Karliner that discrepancies between the E142 neutron data (together with the EMC proton data) and the Bjorken sum rule can be explained by higher-twist contributions. Our conclusion, based on the accurate data available on unpolarized structure functions, is that such contributions are strongly constrained to be negligible, even in the lower energy regions explored by the SLAC deep-inelastic scattering experiments. Thus, based on an analysis of all present polarized data, we re-affirm that the Bjorken system of equations is violated at the two standard-deviation level. Following the spirit of the explanation we have already proposed to this problem, we show that it is sufficient to slightly modify the normalization of the Wilson coefficients involved to obtain a self-consistent and physically sensible description of the nucleon spin structure functions.