Abstract
The structureless exp(1.8t) behavior of recent pp data for |t|2 GeV2, s=53 GeV, is shown to be in sharp disagreement with dip-structure predictions from popular models of diffraction. Flip amplitudes, real parts, and large-angle effects are quantitatively insufficient to resolve the discrepancy. Modifications of some familiar ideas on diffraction (like eikonalization) seem necessary.