Empty Is Emptied: How Frame Choice Affects Previous State Inferences

Abstract

Frame choice generally affects readers’ evaluations: objects or events framed positively are evaluated more positively than objects or events described negatively (Levin and Gaeth 1988; Holleman and Pander Maat 2009). This so-called “valence-consistent shift” is an important rhetorical resource: speakers may steer hearer inferences and evaluations just by choosing a particular attribute frame. Moreover, the framing choice in utterances as this glass is half empty seems to affect inferences about “previous states”, that is about the change leading to the current state of affairs. This chapter replicates and extends the experimental work on these previous state inferences. Furthermore, it explores the mechanisms behind this framing effect. Experiment 1 provides a Dutch replication of the seminal study by Sher and McKenzie (2006) on half full and half empty cups. It confirms the role of prior state reference points in frame interpretation: half full tends to be comprehended as “fuller than before”, while half empty is seen as “emptier than before”. Our experiment 2 constitutes a conceptual replication of the first study, using a variety of written scenarios and a variety of descriptors different from full and empty. Once more, the framed component tends to be taken to have increased over time. In closing, we relate the results to the notion of reference points and to the concept of argumentative orientation. More specifically, we propose that the reference point idea instantiates a more general principle of argumentative orientation.