An fMRI study of processing novel metaphoric sentences

Abstract

Due to inconsistent findings, the role of the two cerebral hemispheres in processing metaphoric language is controversial. The present study examined the possibility that these inconsistent findings may be due, at least partly, to differences in the type (i.e., words vs sentences) or the familiarity of the linguistic material. Previous research has shown that novel two-word metaphoric expressions showed stronger activation in the right homologue of Wernicke's area for the novel metaphors than for both literal expressions and unrelated word pairs. In the present study fMRI was used to identify the left (LH) and the right hemisphere (RH) neural networks associated with processing unfamiliar, novel metaphoric sentences taken from poetry, as compared to those involved in processing familiar literal sentences and unfamiliar nonsensical sentences. Across participants, several left lateralised brain regions showed stronger activation for novel metaphoric sentences than for the nonsensical sentences although both types of sentence represent unfamiliar linguistic expressions. Moreover, the metaphoric sentences elicited more activation in the left dorsolateral prefrontal cortex and the posterior middle temporal gyri than did both the literal sentences and the nonsensical sentences. The increased activation in these brain regions might reflect the enhanced demand on the episodic and semantic memory systems in order to generate de-novo verbal semantic associations. The involvement of the left posterior middle temporal gyri could reflect extra reliance on classical brain structures devoted to sentence comprehension.

Acknowledgements

Support for this work was provided by a grant from US-Israel binational science foundation (BSF, grant # 2003317) to Faust and Jung-Beeman. We would like to thank Chen Kleinman for the useful assistance in preparing the linguistic stimuli.