Illusions of perception of 3-D house models

In five experiments some edgelines on different polyhedrons (house models in 3-D) were, from a certain vantage point, optically confluent, ie in optical prolongation of each other in 2-D on the retinal image and on photos. Other edgelines on the same polyhedrons were non-confluent, ie optically separate in 2-D. These conditions were found to lead to five different illusory shapes in 3-D. Five spatiofigural illusions were perceived. From these findings an edgeline principle is formulated that: "a straight edgeline in 2-D, whether confluent or separate, is perceived as a unitary and continuously straight edgeline in 3-D". To this is added a supplementary perceptual principle, an amodal completion principle. In the experiments reported here, the illusory perception of shapes in 3-D with confluent edgelines as well as the veridical perception of other shapes in 3-D with only separate or non-confluent edgelines could all be explained by the edgeline principle and the amodal completion principle. By applying the concepts of edgeline confluence and the edgeline principle, a new explanation of the Kopfermann (1930 Psychologische Forschung 13 293- 364) cube phenomena is proposed together with one example of how to test this explanation experimentally.