Neural Mechanisms of Figurative Language Processing

An embodied cognition view postulates that concrete words and sentences denoting actions activate motor and premotor cortices (Aziz-Zadeh et al., 2006; Pulvermüller 2012; Desai et al., 2013). However, it remains unclear whether processing of abstract non-literal language relies on sensory-motor activations as well (Cacciari et al., 2010; Raposo et al., 2009; Boulenger et al., 2012; Cacciari and Pesciarelli, 2013). Here, we use bilateral Transcranial Magnetic Stimulation (TMS) and Motor Evoked Potentials (MEPs) to investigate whether motor activation accompanies processing of both literal and figurative language. Russian native speakers read literal, idiomatic, and metaphoric sentences incorporating hand-related action words as well as control abstract sentences, while we delivered single-pulse TMS (1) at the action verb, before figurative or literal meaning is disambiguated and (2) at the end of the sentence to test the sensitivity of cognitive mechanisms of sentence processing and engagement of neural networks to both the word meaning per se and the context. The participants’ MEPs were continuously recorded and MEP differences between conditions and TMS stimulation sites were analysed. The results suggested that motor cortex excitability during literal and figurative sentence processing in the limited context was comparable across all types of stimuli. When broader disambiguating context was provided (i.e. complete condition), motor activation significantly decreased for idiomatic sentences in all conditions and showed a more complex pattern for Literal and Metaphoric ones. This pattern suggested a correlation between the level of motor cortex excitability and the proportion of the basic verb meaning, preserved in the semantics of the verb, which became significant only in the broad context.