An interpretation of masking by light in terms of the sustained-transient dichotomy is presented. It is proposed that presentation of a uniform flash has two effects: a masking effect, which occurs within the transient system, and a facilitation effect due to increased adaptation level. These two components have been separated by the employment of uniform flashes (flicker around a mean luminance) that do not change adaptation level. Results of such studies generally suggest that masking by light occurs only when targets are detected by the transient system. However, not all results are consistent with the sustained-transient model. Reexamination of other recent literature reveals that the motion/pattern dichotomy underlying the model is not tenable. Moreover, other evidence appears to refute the view that the sustained system dominates fovea vision while the transient system is more responsive in the periphery. I suggest a modified version of the dichotomy and propose a simple receptive field model of sustained and transient mechanisms.This article is written with three goals in mind. The first is to propose a new interpretation, based on the sustained-transient model, of the masking-by-light phenomenon. The second is to reexamine the validity of the sustained-transient dichotomy and the criteria on which the dichotomy is based. And the third is to present a receptive field model which seems to account for most of the differences between sustained and transient mechanisms.The sustained-transient model (Breitmeyer & Ganz, 1976;Keesey, 1972;Kulikowski & Tolhurst, 1973;Tolhurst, 1973) has become one of the most influential theories in vision. It has been used to explain a number of phenomena, including metacontrast masking (Breitmeyer & Ganz, 1976) and pattern masking (Legge, 1978;Pantle, 1983). In this paper, I will attempt to show that the model can be used to largely explain masking by light. Most of the evidence for my interpretation comes from the studies I have conducted, although I will also include converging evidence from other sources.My analysis derives largely from studies in which observers view sinusoidal gratings in the presence of uniform flicker. Data generated by this paradigm were generally consistent with an explanation of masking based on the sustained-transient dichotomy.As the studies progressed, however, it became clear that not all of the results could be reconciled with the most common versions of the sustained-transient model. Moreover, a review of other data published

Observers detected drifting gratings presented on a steady uniform field or on a background of uniform flicker with the same time-averaged luminance. Masking effects were obtained whether the flicker mask was presented to the eye on the same side as the test grating or to the one on the opposite side. Unlike masking-by-light studies, in which the mask changes the mean luminance of the display, dichoptic effects were only slightly smaller than those obtained monoptically. Moreover, when observers detected flicker masked by gratings, the dichoptic effect was as large as or larger than the monoptic masking. These results are consistent with the view that masking by light involves separate steady-state luminance and transient components. It is likely that only the transient effect occurs under conditions of dichoptic stimulation.