Conceptual and lexical determinants of nonentrenched thinking

Abstract Two experiments investigating information-processing consequences of entrenched and nonentrenched concepts are reported. An attempt is made to distinguish between these two kinds of concepts by using two variables—the naturalness of the occurrence described by a concept and the familiarity of the name used to refer to that occurrence. In each experiment a given conceptual system was expressed in four alternative forms by crossing concept familiarity (naturalness) with lexical familiarity. The experiments used a concept-selection task in which subjects were required to characterize an event based on a preliminary piece of information and a final, confirmatory piece of information. The results indicated that the locus of nonentrenchment lies in using a familiar name to identify an unfamiliar occurrence or in using an unfamiliar name to identify a familiar occurrence. An information-processing model of task performance provided a very good account of the latency data and scores from the concept-selection task correlated with scores from a set of psychometric reasoning tests. The distinction between entrenched and nonentrenched concepts can be interpreted in terms of interference theory, and it also has implications for the way we think about induction and human intelligence.