Theoretical and simulation approaches to E-region irregularities (gradient drift and Farley-Buneman instabilities) are reviewed, and an account is given of some relevant observations. A new hybrid linear dispersion relation is also derived and presented. The most important problem that cannot be explained by more straightforward theories is the saturation of the phase velocity to the ion acoustic speed (C_s saturation). This phenomenon is well-known from equatorial electrojet radar observations. Recent particle simulations have yielded an interesting new explanation for the (C_s saturation, which has been named flow angle stabilization: the phase velocity is not actually (C_s saturated, but the flow angle distribution of the spatial power spectrum is highly asymmetric. The asymmetry is such that the most intense waves propagate at the k·E < 0 edge of the linearly unstable sector, and thus the phase velocity of the most intense waves is close to (C_s. Depending on the level of larger scale turbulence, the radar observes varying degrees of (C_s saturation. If the larger scale turbulence level is high (equatorial electrojet case), the local flow angle fluctuates, and there are always sub-regions within the scattering volume with local flow angles favourable for the detection of the most intense waves. Under these conditions, the spectra show (C_s saturation. If the larger scale turbulence level is lower, there will not always be enough mixing of the flow angle for even the most intense waves to be observed. In these cases, the mean Doppler shift will be proportional to the electric field, i.e. it will obey the linear theory.