"Don't fence me in!": Examining the source of bias within Müller-Lyer target aiming.

The Müller-Lyer illusion distorts the perceived length of a line shaft based on the relative orientation of tail-ends. This illusion has been heavily adapted to evaluate the visual processes underlying both perception and action. While much of the evidence suggests an influence of this illusion on actions such as manual aiming, the debate surrounding this influence continues. Indeed, it is possible that the tail-ends independently steer extra-retinal inputs regarding target position, which could inadvertently bias any aiming attempts. To this end, participants completed rapid aiming movements toward a series of tail configurations (tails-in, tails-mid, tails-out, tails-combined) either with or without the line shaft; thus nullifying any misperception in length that is typically caused by the illusion (Exp. 1). The findings showed differences in movement displacement that were effectively shifted toward the tails, irrespective of whether there was a shaft present. To extend these findings, we additionally introduced configurations whose tails were oriented left and right (tails-left, tails-right) (Exp. 2). As well as replicating the previously highlighted differences in movement displacement, we also showed a shift in the movement direction toward the tail-ends. Each of these movement effects were isolated to the very end of the movement rather than the early portions. While these findings do not disqualify the influence of illusion configurations on action, they nonetheless indicate a different origin to the relative metrics that are suggested to underlie perception.