Poles and poltergeists in e+e- →D D ¯ data

A recent report of e+e-→DD¯ events by the BESIII collaboration suggests the presence of a structure R at 3900 MeV. We argue that this structure, called G(3900) in the past, is not in fact due to a new cc¯ resonance but rather naturally emerges due to a combination of interference between nearby resonances and the opening of the D∗D¯ channel. We further find that the appearance of this structure does not require suppression because of a radial node in the ψ(4040) wave function, although a node improves fit quality. The measured e+e- coupling of ψ(4040) is found to be substantially smaller than previously estimated. In addition, we report new corrections to the measured cross section σ(e+e-→DD¯) at energies near ψ(3770).