Radar observations of field line resonances (FLRs) show two categories of waves with large or small azimuthal wavenumbers (m). Low m resonances are relatively well understood theoretically, however high m waves appear to be continually driven and display characteristics which can only be understood by considering the effects of the driver. We discuss an heuristic model driving term suggested by [Southwood, 1974>, and recently adopted by [Fenrich and Samson, 1997> in an attempt to explain their observations. We show how their model introduces non-causal (unphysical) features which are not good approximations to reality. We offer a possible alternative heuristic MHD model which can better explain the observational features of high m FLRs. The new model formulation may generate modal solutions with the observed equatorwards phase motion, growth in time, and radially oscillatory characteristics. ¿ 1998 American Geophysical Union