A growing body of experimental work, including a number of the articles cited above, suggests that the observation of the effects of a gravity field apparently acting on people and objects in a distant scene can provide a significant visual “vertical” reference (Howard,
1982). We employ the term “pictorial gravity” (or “visible gravity”) to distinguish the apparent gravity from physical gravity. Pictorial gravity is not always aligned spatially with physical gravity, as it happens, for instance, when we watch a tilted picture or movie. Moreover, while the magnitude of physical gravitational acceleration of an object is constant, the acceleration of the resulting retinal image is inversely related to the apparent viewing distance (assuming an object's motion in a fronto-parallel plane). However, pictorial cues (e.g., familiar size of people and objects, linear perspective, shading, texture gradient) built-in to the visual scene may help viewers gauge the spatial scale and orientation of the scene (Bingham,
1993) and estimate the effects of pictorial gravity on the objects of that scene (Miller et al.,
2008). Thus, when pictures of static postures of a human standing on a platform and tilted in the roll plane are presented in different orientations, observers are able to judge the stability of this human body relative to pictorial gravity even for directions that are not concordant with the direction of physical gravity (Lopez, Bachofner, Mercier, & Blanke,
2009). In addition, it has been shown that the discrimination precision of a target motion duration is higher when the target accelerates in the downward direction of a virtual room (consistent with pictorial gravity) than when it accelerates in the upward direction (violating gravity), and this is so whether the virtual room and target motion direction are aligned with the observer or they are tilted by 45° (Moscatelli & Lacquaniti,
2011). Finally, it has been shown that viewing a rotated photograph or video clip with strong polarization cues, which indicate relative “up” and “down” directions in the picture, can alter the perceived direction of absolute “up” and “down” directions in the world (Dyde, Jenkin, & Harris,
2006; Jenkin, Jenkin, Dyde, & Harris,
2004; Jenkin, Dyde, Jenkin, Zacher, & Harris,
2011). Thus, perceptual biases have been documented in relation to the orientation of the stimuli relative to gravity (Chang, Harris, & Troje,
2010; Lobmaier & Mast,
2007; Lopez, Bachofner, Mercier, & Blanke,
2009) and to intrinsic visual references (Jenkin, Jenkin, Dyde, & Harris,
2004; Tadin, Lappin, Blake, & Grossman,
2002).