Irony: Grade Salience and Indirect Negation

Abstract

According to the graded salience hypothesis, salient (e.g., conventional, frequent, familiar) meanings should be activated before less salient meanings are retrieved (R. Giora, 1997). The graded salience hypothesis thus predicts that irony should take longer to understand than non-ironic language, because its comprehension should involve activation of the salient, literal meaning initially, before the less salient ironic meaning is derived. According to the indirect negation view of irony (R. Giora, 1995), irony comprehension should involve retention of the activated literal meaning so that the comprehender may compute the difference between the (usually desirable) state of affairs alluded to by the literal meaning and the less desirable, ironicized situation. Evidence from three experiments supports the graded salience hypothesis and the indirect negation view of irony. Experiment 1 showed that target sentences took longer to read in ironically than in literally biasing contexts. In Experiments 2 and 3, response times to test words that were either literally or ironically related to the target sentences were measured. Both literal and ironic target sentences facilitated only the literal meaning 150 and 1,000 msec after their offset. However, 2,000 msec after offset of the ironic target sentences, the ironic meaning became available and the literal meaning was still as active.