Abstract

Several recent writers have argued that the rejection of hypothetical constructs is one of the defining features of radical behaviorism. The present discussion argues that this claim is ill-founded and based on an erroneous distinction regarding different kinds of theoretical constructs. All constructs, including those commonly employed by behavior analysts, are argued to be inherently hypothetical, because they provide a causal basis for extending empirical findings to new sets of variables. Moreover, the constructs employed by radical behaviorists are not qualitatively different from those often employed by cognitivists . The result is that much of the basis for recent criticisms of cognitive psychology by behaviorists disappears. To the extent that differences do remain it is not in the qualitative nature of the two enterprises, but instead in attitudes about parsimony in the relation between data and theory