Distinction between the literal and intended meanings of sentences: a functional magnetic resonance imaging study of metaphor and sarcasm.

To comprehend figurative utterances such as metaphor or sarcasm, a listener must both judge the literal meaning of the statement and infer the speaker's intended meaning (mentalizing; Amodio and Frith, 2006). To delineate the neural substrates of pragmatic comprehension, we conducted functional magnetic resonance imaging (fMRI) with 20 normal adult volunteers. Participants read short stories followed by a target sentence. Depending on the context provided by the preceding stories, the target sentences were classified as follows: (1) metaphor versus literally coherent; (2) metaphor versus literally incoherent; (3) sarcasm versus literally coherent; and (4) sarcasm versus literally incoherent. For each task pair, we directly compared the activations evoked by the same target sentences in the different contexts. The contrast images were incorporated into a 2 (metaphor and sarcasm)×2 (literal coherency and incoherency) design. Metaphor-specific activation was found in the head of the caudate, which might be involved in associating statements with potential meanings, and restricting sentence meanings within a set of possible candidates for what the speaker intended. Sarcasm-specific activation was found in the left amygdala, which is an important component of the neural substrates of social behavior. Conjunction analysis revealed that both metaphor and sarcasm activated the anterior rostral medial frontal cortex (arMFC), which is a key node of mentalizing. A distinct literal coherency effect was found in the orbital MFC, which is thought to be involved in monitoring. These mesial frontal areas are jointly involved in monitoring literal coherency and mentalizing within social contexts in order to comprehend the pragmatic meanings of utterances.