Reprints and permission:

People perceive the world as being stable and are capable of making accurate, visually guided movements. This is remark-able considering the fact that visual perception relies on input from the retina, which, because of eye movements, is unstable. The apparent conflict between the lack of stability in visual input and the stable nature of conscious visual perception is often referred to as the problem of visual stability. An intuitively appealing solution to this problem is to assume that only early visual areas are affected by eye move-ments. It is an old idea, often attributed to Helmholtz or even pre-Socratic philosophers, that information about eye position is crucial in maintaining visual stability (Grusser, 1986). The spatiotopic hypothesis posits that action and conscious percep-tion are based on a world-centered (spatiotopic) representa-tion, which is constructed by combining gaze-centered (retinotopic) maps with eye-position information. However, there is little empirical support for this hypothesis. The most convincing evidence so far has been provided by a functional magnetic resonance imaging (fMRI) study showing spatio-topic responses in the human middle temporal cortex (d’Avossa et al., 2007), but this finding has recently been contested (Gardner, Merriam, Movshon, & Heeger, 2008). The evidence is somewhat different for multimodal parietal areas, where gaze-modulated visual receptive fields are frequently observe