Abstract
Physical connectedness between objects is one of the factors that determines their representational unity. Previous studies, in which connectedness of two-dimensional (2-D) shapes was used to manipulate representational unity, have shown speeded recognition for connected shapes compared to disconnected shapes. In this study, a distinction is made between 2-D and three-dimensional (3-D) connectedness. The first refers to the connectedness in the image itself, while the latter refers to the connectedness of the interpreted objects (that is, the objects that gave rise to the image). A projection of a pair of 3-D objects can show these objects to be connected in the image while they can be separated in 3-D space (e.g. one object is in front of another and partly occludes it). We were interested in projections of 3-D objects in which there is either congruency or incongruency between 2-D and 3-D connectedness. In two experiments, matching tasks were used to tease apart the 2-D and 3-D levels of connectedness. Two projections of 3-D objects were presented simultaneously. These projections could have the same orientation, or they could be 180í rotated versions of each other. Participants had to judge whether the two projections were identical to each other or not, regardless of their orientation. It was found that 3-D connectedness facilitates object transformation (such as mental rotation) compared to 2-D connectedness. From the results of the experiments it appears that the human visual system relies on the interpretation of the image, rather than on the image itself.