Neural basis of idiomatic language comprehension

The comprehension of idiomatic language is a matter of debate for what concern its neural basis and the psycholinguistic theories that better describe the process. Neuropsychological and neuroimaging studies report highly different outcome, especially about the specific role of the two cerebral hemispheres: while some studies show the involvement of th e left inferior frontal gyrus (LIFG), other data support the crucial role of right-hemispheric regions, and particularly of the middle/superior temporal area. Different psycholinguistic models have been proposed to explain how we can understand these semantically eccentric expressions; these models can be broadly divided in those that predict direct access to the idiomatic meaning and those in which it is dependent on the suppression of its literal meaning. Fifteen volunteers took part to the present experiment. They silently read 360 idiomatic and literal Italian sentences and decided whether they were semantically related or unrelated to a following target word, while their EEGs were recorded from 128 electrodes. Only purely idiomatic expression were selected, excluding metaphors or fixed expressions in general. Word length, word frequency, sentence comprehensibility and familiarity, sentence cloze probability, overall salience and abstractedness were balanced. Subjects were faster in responding to literal than idiomatic sentences, probably suggesting a difference in task difficulty. ERP components elicited by sentence final words showed the existence of at least three processing stages in which differences were observed between ERP data elicited by idiomatic vs. literal sentences. N2 component evidenced a greater amplitude in response to idioms between 250 and 300 ms of latency. swLORETA for this effect revealed a generator in the left fusiform gyrus. It was also found a bilateral N400 component between 360 and 550 ms, which was much larger to idioms than literal sentences. The intra-cortical generators for this effect included the left and right FGs, the left cingulate gyrus, the right limbic area, the left superior temporal gyrus and left frontal inferior gyrus. The activation of left and right limbic regions suggests that they have a role in the emotional connotation of colorful idiomatic language. Finally an anterior P300 was identifie d between 600 and 800 ms, that was more positive to idiomatic than literal sentences. Overall, the data indicate a bilateral involvement of both hemispheres during idioms comprehension, with a special role for right hemisphere, especially after 350 ms post-stimulus. They also support the view of a direct access to the idiomatic meaning of figurative language, not dependent on the suppression of its literal meaning. ERPs synchronized with the target word showed a larger right centroparietal N400 to associated than non-associated targets (not differing as a function of sentence type), and a greater right frontal P600 to idiomatic than literal associated targets.