Single-spin asymmetries and Qiu-Sterman effect(s)

I discuss the relation between the Qiu-Sterman effects on one hand and the Collins, Sivers and Boer-Mulders effects on the other hand. It was suggested before that some of these effects are in fact the same, thus providing interesting connections between transverse-momentum dependent twist-2 functions and collinear twist-3 functions. Here I propose an alternative way to reach similar conclusions.