Using Resource Depletion to Examine Processing of Familiar and Unfamiliar Metaphors

An important question in research on metaphor comprehension is whether familiar and unfamiliar metaphors are understood through separate processes. In prior research, the Categorization model (Glucksberg, 2003) specifies that comprehension of familiar and unfamiliar metaphors relies on the same set of controlled processes, whereas the Career of Metaphor model (Bowdle & Gentner, 2005) specifies separate processes: a controlled process for unfamiliar metaphor comprehension and an automatic process for familiar metaphor comprehension. The present study assessed these competing models using a resource depletion manipulation. Resource depletion (Muraven & Baumeister, 2000) refers to the gradual depletion of a limited executive or cognitive control resource during activities that draw upon this resource, eventually resulting in temporary performance impairments on any activity that requires the exertion of cognitive control. Because depletion of the cognitive control resource negatively impacts higher-order controlled processes but does not affect automatic processes, the present study aimed to compare the effects of resource depletion on comprehension of low- and high-familiarity metaphors to reveal whether automatic or controlled processes are recruited. In two experiments, resource depletion was induced using a modified Stroop task, followed by a different metaphor-related task in each experiment. In this task, participants in both experiments were presented with the same set of short literal or metaphor (high, low, and mid-familiarity) sentences. In Experiment 1, participants were instructed to indicate when they had understood each sentence. In Experiment 2, participants were instructed to categorize the meaning of each sentence as “concrete” or “abstract”. Results for Experiment 1 indicated that resource depletion slowed comprehension by much greater margins on mid- and low-familiarity metaphors compared to highly familiar metaphors. Results for Experiment 2 were inconclusive, as performance of the concrete/abstract categorization task was not sensitive to resource depletion. Overall, the results support elements of both the Categorization and Career of Metaphor models, but support appears to be stronger for the Career of Metaphor model.