Abstract
It is proposed that much human inference is basically analogical, and the implications of this theory for cognitive development are examined. Analogy entails mapping the problem representation (target) into a structurally similar schema (base), which has been learned through experience. It can help explain some significant reasoning phenomena, such as errors in the four-card problem. The theory is explicated by analysing the processes entailed in transitive inference and class inclusion tasks. Two sources of difficulty that arise in analogical reasoning are (a) mapping problems to inappropriate schemas, and (b) processing loads imposed by mappings. The loads are greater for more complex concepts. A conceptual complexity metric is defined, which distinguishes four levels of structure mapping. The levels increase in flexibility and abstractness, but at the price of increased processing load. The relation of the model to contemporary accounts of reasoning is examined, and it is shown to be consistent with the theory of pragmatic reasoning schemas, and other knowledge-dependent approaches, while maintaining an important role for structure. Implications of the theory for the stage debate and other issues in cognitive development are considered.