Abstract
Studying human judgements of a moving object's position is complicated by the fact that the position is constantly changing, so that one has to specify when one wants to know the position. If a flash is used to indicate the time, the moving object is judged to be at a position that it only reaches some time after the flash. We presume that this is because the brain has no special mechanism for determining simultaneity, so that the position of interest has to be ‘sampled’ at the indicated moment. Since this takes time, and can only be initiated after the flash (or some other signal), an asymmetry arises, whereby the moving object's position is judged too late. To test this account we presented objects that passed slightly above or below a fixation point as they moved from left to right across a screen. We asked subjects to indicate the moving object's position at the moment that the fixation point flashed, or at the moment that they heard a tone. We varied the object's velocity and where it was at the moment of the flash or tone. Subjects mislocalised a slowly moving object's positions in the direction of motion, but as the object's velocity increased this mislocalisation was overshadowed by a tendency to see the moving object close to where it passed the fixation point (irrespective of whether the signal was a flash or a tone). A simple four-parameter model (sampling delay, variability in sampling time, spatial precision and central bias) describes the data reasonably well. This model is not only consistent with various forms of flash-lag, but it is also consistent with the perisaccadic compression of perceived space when there are visible references.