Abstract
Two experiments are described, the purpose of which was to investigate the presence of a misalignment illusion caused by Poggendorff-like conditions in two paintings by Peter Paul Rubens, both depicting the Descent from the Cross, one located in Antwerp (Belgium), the other in Lille (France). The first shows a geometrical misalignment made by Rubens in a minor detail, which is considered proof that the artist observed the Poggendorff illusion. The second painting, instead, shows a perfect geometrical alignment in a similar detail. In experiment 1, participants were asked to align a top segment to a lower one in two types of stimuli: a full-size digitally manipulated reproduction of the painting and a Poggendorff-like configuration that recalled the painting's lines displacement and tilt. Adjustments were performed from two distances, one up close (painting distance) and one from below and far (observation distance). Results confirmed the presence of the Poggendorff illusion, but mean adjustments significantly differed from the misalignment perpetrated by Rubens. Experiment 2 was set up in a similar fashion with the Lille painting. Results confirmed the presence of the Poggendorff illusion also in this painting; however, the alignment by Rubens coincides with the geometrical one. Results from both experiments do not support the claim that Rubens observed the Poggendorff illusion and therefore corrected for it in the Antwerp painting. An alternative account is discussed, which relates to the structural layout of the painting.