Abstract
We discuss some of the theoretical arguments for the existence in Compton scattering of right-signatured fixed poles with polynomial residues. We show that if one could "switch off" the strong interactions, then a fixed pole with residue linear in q2 (the photon mass squared) would be necessary for the consistency of the fixed-q2 dispersion relation for νT2 (whose absorptive part νW2 is measured in inelastic electroproduction). We show that if the above conjecture is correct, then there must be some energy dependence in νW2 over and above the conventional leading Regge form (Pomeranchukon plus fA2). Evidence is presented for the presence of such "nonleading behavior" in a similar process. In addition we show why the on-shell σtot(γp) could be compatible with the neglect of such a nonleading term. We find that a fixed pole with polynomial residue and the correct q20 Thomson limit can be accommodated by the present data on νW2 at large q2. With the above assumptions on the fixed-pole behavior, we predict the high-energy behavior of νW2 and find that asymptotically it must fall to a value substantially less than its present maximum magnitude.