Science and politics: Dangerous liaisons

In contrast to the opinion of numerous authors (e.g. R. Rudner, P. Kitcher, L. R. Graham, M. Dummett, N. Chomsky, R. Lewontin, etc.) it is argued here that the formation of opinion in science should be greatly insulated from political considerations. Special attention is devoted to the view that methodological standards for evaluation of scientific theories ought to vary according to the envisaged political uses of these theories.I wish to thank the Alexander von Humboldt Foundation for having supported the work on this paper.     

The Craik-O’Brien-Cornsweet effect: New varieties and their theoretical implications

New visual displays related to the Craik-O’Brien-Cornsweet effect are presented and their relevance for theories of the effect is assessed. The displays include a luminance cusp that does not induce the standard effect, an illusory chessboard, a brightness variant of the impossible staircase, and a display in which brightness varies with stimulus motion. Theories of the effect are grouped within the nonisomorphistic, the cognitive, and the mechanistic approaches. The nonisomorphistic approach is criticized as being seriously incomplete. Existing cognitively oriented proposals are criticized as inadequate. Mechanistic theories are divided into integration theories, which model the visual processing stages with standard mathematical operations such as differentiation and integration, and filling-in theories, which stress the processing dynamics of interconnected neural networks. The reported variations of the Craik-O’Brien-Cornsweet effect support the latter class of theories.