Visual illusion and line bisection: a bias hypothesis revisited.

It is generally accepted that visual illusions affect line bisection in the predicted direction. However, it has been reported an illusionary bias which seems questioning such general view. In a previous study, participants bisected lines flanked at both ends by two pairs of arrows, pointing in the same direction. The medialmost vertices of one pair converged on the line (converging arrows), whereas those of the other pair did not (non-converging arrows). Participants bisected lines toward the base of the arrows, i.e., toward the wider end of the stimulus and in the direction opposite to that predicted by the Baldwin illusion. However, the bisection bias was also directed away from the location of the converging arrows. We investigated what is the main factor affecting line bisection: arrows orientation, as previously suggested, or interference effects related to the location of converging arrows. In experiment 1, participants bisected lines flanked by converging versus non-converging arrows. Results confirmed the presence of a bisection bias directed not only toward the base of the converging arrows but also away from their location. In experiment 2, the arrows were located more internally, so that their medialmost vertices always converged on the line. Results showed that the bisection bias was directed away from the location of the arrows regardless of their orientation. It is suggested that the previously reported bisection bias did not depend on arrows orientation, but rather on interference effects related to converging arrows position. The theoretical implications of the results are discussed.