To test the Silence hypothesis (1970) that subjects in a Right-blank (Rb) reinforcement combination treat blank as right, an experiment similar to that of Silence was performed. Although the results in the Wrong-blank (Wb) and Right-Wrong (RW) groups were similar to the Silence data, those in the Rb group were different: nearly half of the Rb subjects treated blank as right and the other half of them treated it as wrong. The latter findings suggest that the subjects in the Rb group who perceived right as being given contingently to their correct responses treated blank as wrong but that the subjects who perceived right as being given noncontingently to their responses treated blank as right.