Is the Concreteness Effect a Result of Underlying Network Interconnectivity?

Two general classifications of theoretical explanations for concreteness effects are evaluated. Process theories are those that emphasize the nature of the processing initiated by the study items, and structural explanations are those that emphasize the role of long-term memory representations of the study items. A series of four experiments is presented in which structural variables such as the number of associates a word has in long-term memory (set size) and the density of the connections among those associates (connectivity) are crossed with concreteness under several different instructional manipulations in the cued recall task. The results consistently demonstrate that concreteness effects and effects related to the associative structure of the study items are not dependent upon one another, but instead have additive effects on performance. The results allow for the elimination of the structural explanations, in favor of one of the process theories: dual code theory. The results are also consistent with the overall framework of memory proposed by the PIER model.