The demise of the identity hypothesis and the insufficiency and nonnecessity of contour relatability in predicting object interpolation: comment on Kellman, Garrigan, and Shipley (2005).

P. J. Kellman, P. Garrigan, and T. F. Shipley's theory of 3-dimensional object interpolation asserts that existing data, as well as logical considerations, support the view that an identical contour interpolation process underlies the interpolation of partially camouflaged and partially occluded objects (modal completion and amodal completion, respectively). Here, the author argues that recent data show that this theory is incorrect and that the logical arguments offered in support of the identity hypothesis depend on specific unverified models of the phenomena in question. Alternative explanations of these effects are developed to show that such phenomena do not logically implicate an identity hypothesis and, in some cases, provide strong evidence against the identity hypothesis. Finally, the author describes several completion phenomena that reveal that the relatability criteria embodied in Kellman et al.'s model are neither necessary nor sufficient for understanding the interpolation processes the model was designed to explain.