Abstract
Illusory line motion (ILM) is an illusion of motion perceived within a line even though the line appears all at once. A canonical demonstration of this phenomenon is as follows. Two squares are presented on a computer monitor. After a brief delay, one of the squares is cued (e.g., flashed) briefly, and a line is presented connecting the two squares. Although the line is presented instantaneously, most observers perceive it as propagating out from the cued square in the direction of the uncued square. Other factors being equal, ILM tends to originate from the square closest to the focus of visuospatial attention. It is commonly observed that visuospatial attention tends to track moving stimuli. We wondered whether attention would track ILM as if it were real motion. To address this question we recorded visual evoked potentials (VEPs) in response to probes flashed at either end of the line in a typical ILM display. The notion was that the N1 component of the VEP would be amplified when the probes appeared at attended locations. There were three conditions: “ILM,” “No-ILM,” and “Object Control.” In the ILM condition we compared the N1 amplitude evoked by probes at the cued and uncued squares presented 400-800 ms following the presentation of the line. The No-ILM condition was similar except that the line was not presented prior to the probe, so there was no illusion of motion. In the Object-Control condition the line was present throughout the trial, so there was again no illusion. We predicted that the N1 amplitude would be greater at the cued square in the No-ILM and Object Control conditions, since attention would be captured by the cue. We further predicted that ILM would draw attention away from the cued square towards the uncued square, which would in turn result in enhanced N1 at the uncued square. Preliminary data uphold both of these predictions. This supports our conjecture that ILM “drags” attention in the direction of illusory motion.