The cyclopean illusion unleashed

Abstract

The cyclopean illusion is the apparent lateral shift of stationary stimuli on a visual axis that occurs when vergence changes. This illusion is predictable from the rules of visual direction. There are three stimulus situations reported in the literature, however, in which the illusion does not occur. In the three experiments reported here we examine those stimulus situations. Experiment 1 showed that an afterimage seen on a stimulus moving on the visual axis does not produce the illusion as reported in the literature but an afterimage seen on a screen does. Experiment 2 showed that the illusion occurs for an intermittently presented stimulus in contrast to what has been reported previously. Experiment 3 showed that a monocular stimulus presented against a random-dot background produced the illusion, also in contrast to what has been reported. The results were consistent with the rules of visual direction.

Two historical strands in studying visual direction1

“…The lack of apparent movement, however, can be explained by (a) a large stimulus (in this case the random dot stimulus) tends to appear stationary and (b) the distance of his far stimuli (30 cm) and near stimulus (15 cm) produces small accommodative vergence movements. For experimental results gathered to counter directly Erkelens's findings and Erkelens and van Ee's conclusions, see Ono, Mapp, and Howard (2002), Khokhotva, Ono, and Mapp (2005) and Ono, Mapp, and Mizushina (2007). For other reports in the literature that show the predicted apparent movement, albeit for simple stimuli as shown in Figure 7, see Carpenter (1988), Enright (1988), Helmholtz (1910/1925), Hering (1879), and Wells (1792).…”

“…The illustration is drawn as though the convergence were completely coupled with accommodation, but when the two stimuli are very close to the observer the occluded eye deviates from the indicated positions in the figure and the predicted extent of the apparent movement would be smaller. (Adapted from Ono et al., 2007.)…”

Binocular Vision: Defining the Historical Directions

“…To check, for example, whether a ruler is straight or whether three nails on a board are in line, one would use one eye, and a judgment can easily be made without attending (5) Nakamizo et al (2008) used a set of vertical parallel lines for their stimulus, but not all the monocular stimuli presented followed the predictions of Wells (1792). The predictions failed when a monocular line was embedded in (Domini and Braunstein 2001;Erkelens andvan Ee 1997a, 1997b;Ono and Mapp 1995;Shimono and Wade 2002;Shimono et al 1998Shimono et al , 2005Shimono et al , 2007 or when the monocular stimuli were placed in front of a monocular random-dot field (Erkelens 2000;Ono et al 2007). For a discussion of these failures, see Mapp et al (in press).…”

“…It could be that neglecting the somewhat subtle distinction between headcentric and relative direction is responsible for the persistence of old ideas based on optics and projection. There are current claims that monocularly seen objects are veridically localised (Erkelens 2000;Erkelens and van de Grind 1994;Erkelens and van Ee 2002;Erkelens et al 1996;Khan and Crawford 2001;cf Mapp et al 2007;Ono et al 2002Ono et al , 2007. All these researchers who claimed the veridicality used relative visual direction tasks, and therefore inferences about the origin of visual direction, an eye, or the cyclopean eye, cannot be made.…”