Are word meanings atomic or complex? An investigation of conceptual knowledge activation in context

While some studies have suggested that conceptual knowledge can be activated selectively (e.g. Potter & Faulconer, 1979; Springer & Murphy, 1992), others have suggested that the same default set of conceptual features is activated, regardless of the context in which a word is used (e.g. McElree, Murphy & Ochoa, 2006; Swinney, Love, Walenski & Smith, 2007). This study investigated whether context-inappropriate features are brought to mind when people interpret modified concepts (e.g. nursery bear; paper boat; birdcage door). In the contextual condition participants evaluated the weight of a modified concept, while in the compositional condition they first evaluated the weight of the unmodified concept before evaluating that of the modified concept. The results revealed that items involving a mismatch between modified and unmodified weight were more difficult to interpret in the compositional condition but not in the contextual condition. These findings indicate that words are interpreted differently in context than in isolation; specifically, they imply that the activation of context-inappropriate features is avoided when words are interpreted in context