On the importance of Lorentz structure in the parton model: target mass corrections, transverse momentum dependence, positivity bounds

We show that respecting the underlying Lorentz structure in the parton model has very strong consequences. Failure to insist on the correct Lorentz covariance is responsible for the existence of contradictory results in the literature for the polarized structure function g_2(x), whereas with the correct imposition we are able to derive the Wandzura-Wilczek relation for g_2(x) and the target-mass corrections for polarized deep inelastic scattering without recourse to the operator product expansion. We comment briefly on the problem of threshold behaviour in the presence of target-mass corrections. Careful attention to the Lorentz structure has also profound implications for the structure of the transverse momentum dependent parton densities often used in parton model treatments of hadron production, allowing the k_T dependence to be derived explicitly. It also leads to stronger positivity and Soffer-type bounds than usually utilized for the collinear densities.