The visual input is influenced by self-motion in a number of qualitatively distinct ways. In our projected work, we will focus mainly on high-amplitude, directed self-motion, rather than on the low-amplitude, possibly random motion (such as fixational eye movements). We will be interested in the effects of saccades and smooth self-motion. Saccades are rapid, ballistic movements of the eye, which shift the image on the retina from one location to another. Smooth, global motion of the image on the retina results from eye or body movements; of particular behavioral interest is smooth pursuit motion, in which the eye rotates so as to stabilize a moving object on the retina.There are a number of technical and conceptual challenges to the study of visual processing in the presence of self-motion, which range from the design of the stimulus through recording of neuronal activity to the analysis of this activity. Saccadic motion cannot be simulated with standard stimulation devices. Different cell types in the retina come with different physiological properties and, in particular, different sensitivities to motion. In the presence of self-motion, we thus have to be able to examine processing in different cell types separately. In primates, smooth pursuit motion stabilizes a moving object on the fovea. Thus, it is important to have access to the activity of foveal cells, in which visual processing may be organized differently from processing in the periphery. Finally, global motion stimulates a large population of cells in the retina, and, hence, it is necessary to have the technological means to record from such large populations and to have the computational methods to analyze their outputs.We propose to investigate how identified types of ganglion cells, the output cells of the retina, represent visual information individually and collectively in the presence of self-motion, in non-human primate and human retinæ. We will combine expertise in engineering, biomedicine, neurobiology, and theoretical neuroscience to develop methods for adequately stimulating the retina and for recording the spiking activity from large populations of genetically and functionally identified ganglion cells. The use of a large microelectrode array to record simultaneously from the fovea and the periphery, together with computational analyses of the recorded activity, will allow us to characterize foveal and peripheral processing and their interaction.Our projected work will be innovative along several directions. First, it will provide the first description of the responses of identified human retinal ganglion cells. Second, it will yield insights about visual computations by neurons, through the analysis of the activity of a large fraction of the retina submitted to self-motion. Third, it will put forth a first paradigm for the study of the interaction between fovea and periphery in non-human primate and human retinæ. While the study of vision in the presence of self-motion has a long history in psychophysics, comparatively little is known on processing in neural circuits. Our proposed investigations will contribute to filling this gap.