Decomposing metaphor processing at the cognitive and neural level through functional magnetic resonance imaging

Abstract

Prior neuroimaging studies on metaphor comprehension have tended to focus on the role of the right hemisphere, without reaching consensus and leaving aside the functional architecture of this process. The present work aimed to break down metaphor comprehension into its functional components. The study rationale is two-fold: on the one hand, the large-scale network model as emerging in cognitive neuroscience led us to a consideration of metaphor as supported by a distributed and bilateral network; on the other hand, we based on the accounts of figurative language put forward in pragmatics and cognitive science to postulate a decomposition of such a network into multiple sub-systems. During scanning, participants implicitly processed metaphorical (familiar and unfamiliar) and non-metaphorical passages, while being explicitly involved in an adjective matching task to be performed after reading the target passages. Several regions showed greater activity to metaphors as compared to non-metaphors, including left and right inferior frontal gyrus, right superior temporal gyrus, left angular gyrus, and anterior cingulate. This pattern of activations, markedly bilateral, can be decomposed into circumscribed functional sub-systems mediating different aspects of metaphor resolution, as foreseen in the pragmatic and cognitive literature: (a) the conceptual/pragmatic machinery in the bilateral inferior frontal gyrus and in the left angular gyrus, which supports the integration of linguistic material and world knowledge in context; (b) the attentional component in the anterior cingulate and prefrontal areas, which is set to monitor and filter for the relevant aspects of context and for the appropriate meanings; (c) the Theory of Mind system along the right superior temporal sulcus, which deals with the recognition of speakers' communicative intentions and is more extensively activated by unfamiliar metaphors. The results have several implications for the field of neuropragmatics, especially on the neuropsychological side and on the right hemisphere hypothesis.