Introduction

Direction selectivity by retinal ganglion cells—the ability of some cells to report the direction of stimulus movement—was described in classic experiments by Barlow and Levick (Barlow et al., 1964; Barlow and Levick, 1965) more than 35 years ago. Despite much theory and many experiments, the mechanism remains unknown. Indeed, it can be argued, with some embarrassment, that small progress has been made since Barlow and Levick's classic studies. Those studies and subsequent ones have recently been the subject of a careful review by four experienced students of direction selectivity, including its codiscoverer (Vaney et al., 1999). Here I will: (1) outline the cardinal features of direction selectivity—that is, those experimental findings that seem universally accepted; (2) discuss some of the requirements for a mechanistic explanation of it; and (3) comment on certain controversies that have sprung up in the past few years.

Retinal direction selectivity appears deceptively simple. Perhaps for that reason, even sophisticated neurobiologists have sometimes offered theories of its mechanism that could have been seen with a few minutes' careful thought to be unworkable. For that reason, it is important to begin with a statement of the fundamental facts.