Abstract
Recent data on π± photoproduction at 3.4 GeV are used to investigate the nature of the amplitudes contributing to these processes. It is found that a large amplitude corresponding to natural-parity G=+1 exchange is necessary to explain the data, even in the region t0.6 GeV2 where Reggeized ρ exchange must vanish. This implies a large contribution from some ρ trajectory or from cuts, absorption, etc. The large G=+1 exchange amplitude needed to explain the ππ+ ratio is only marginally consistent with the vector-dominance model.