For scalar light, the current is the familiar expectation value of the intensity-weighted momentum operator. Current is distinct from the local wavevector (not weighted) that could be observed by means of quantum weak measurement. For vector light, the current (Poynting vector) contains an additional term corresponding to the photon spin, recently identified for paraxial light by Bekshaev and Soskin but valid generally after a modification to restore 'electric–magnetic democracy'; this term has physical consequences. A number of examples demonstrate that there is usually no connection between optical vortices and angular momentum, and between C singularities and angular momentum. The optical wave current is distinct from the rays of geometrical optics in all nontrivial cases.