Any of a variety of hypotheses entertained by subjects in experiments on verbal conditioning that attributes reward to some form of behavior other than that being reinforced, but leads to reward more often than would happen by chance because the behavior is somehow related to what is being rewarded.