Commentary on Kendler'S Paper: An Alternative Perspective

Abstract

Kendler's theory represents an important step in researchers' attempt to understand the ontogeny of human learning. If the operational definitions offered for (and assumptions concerning) incremental learning and hypothesis testing is accepted, the data are particularly compelling. Kendler's central thesis is that (a) the absence of win-stay responding to a specified stimulus attribute implies the presence of associative (incremental) learning and (b) hypothesis-testing behavior always implies win-stay responding to a specified attribute. There is, however, some evidence in the developmental literature that contraindicates both assumptions. The stochastic version of hypothesis theory explored by Kendler is, of course, congenial to her assumptions. Other versions of the theory, however, particularly those concerned with the behavior of children and nonhuman primates, question their validity. This chapter demonstrates that when an expanding body of research (and theory) is considered, the basic assumptions of Kendler's revised mediation theory encounter some difficulties.