Metaphor pervades discussions of important socio-politicaltopics. Recent research indicates that metaphorical languagecan influence how people reason about such topics, potentiallyaffecting real-world decision-making. In this study, we reporton research into the effects of metaphor on evaluative judg-ment, another aspect of decision-making that has been lesswell studied than reasoning. We use a cross-linguistic differ-ence in the metaphors used by English and Spanish speakersto discuss economic change to investigate how metaphoricallanguage affects evaluative judgment. We show that the imageschematic information inherent in the semantics of the differ-ent metaphors performs a central role in shaping this process.

In this study we investigate how lexical-semantic relations as-sociated with the literal meaning (and abstract meaning) arebeing accessed across the brain during familiar metaphor com-prehension. We utilize a data-driven whole-brain searchlightsimilarity-decoding analysis. We contrast decoding metaphoricphrases (”she’s grasping the idea”) using distributional seman-tic models of the verb in the phrase (VERB model) versus thatof the more abstract verb-sense (PARAPHRASE VERB model)obtained from literal paraphrases of the metaphoric phrases(”she’s understanding the idea”). We showed successful decod-ing with the VERB model across frontal, temporal and parietallobes mainly within areas of the language and default-modenetworks. In contrast, decoding with the PARAPHRASE VERBmodel was restricted to frontal-temporal lobes within areas ofthe language-network which overlapped to some extent withsignificant decoding with the VERB model. Overall, the re-sults suggest that lexical-semantic relations closely associatedwith the abstract meaning in metaphor processing are largelylocalized to language and amodal (multimodal) semantic mem-ory systems of the brain, while those more associated withthe literal meaning are processed across a distributed seman-tic network including areas implicated in mental imagery andsocial-cognition.

Metaphor is an important part of everyday thought and language. Although we are often not aware of metaphor in everyday speech, on occasion, a particularly creative or novel use of metaphor will make us pay attention. It has been hypothesized that one of the driving cognitive factors behind the use of novel metaphor is a need to describe a new reality (as opposed to a preexisting reality) that would otherwise be difficult to convey using conventionalized metaphor. To this extent, novel metaphor use in everyday language may be more associated with episodic memory demands in contrast to conventional metaphor that is associated with semantic memory. To test this idea we analyzed novel metaphor use in the Hippocorpus --- a corpus of more than 5000 recalled and imagined stories about memorable life events in the first person perspective. In this dataset, recalled events have been shown to rely on episodic memory to a greater extent than descriptions of imagined events (i.e., narrating an event as if it happened to you but not describing an event that actually happened to you), which largely draw on semantic memory. We hypothesized that novel metaphor use during event narration should be modulated by the extent to which language users are able to draw on primary experience to describe events. We found that novel metaphor counts in recalled events were significantly higher than imagined events. Importantly, we found that factors that influence the extent to which language users are able to draw on primary experience during event narration (i.e., openness to experience, similarity to one's own experience, and how memorable or important an event was) modulated novel metaphor use in different ways in imagined compared to recalled events. The work paves the way for using large scale corpora to analyze underlying cognitive processes that modulate metaphorical language use.

Recent work shows that distributional semantic models can be used to decode patterns of brain activity associated withindividual words and sentence meanings. However, it is yet unclear to what extent such models can be used to study anddecode brain activity patterns associated with specific aspects of semantic composition such as the negation function. Inthis paper, we investigate the extent to which distributional semantic models of action-verbs correlate with brain activityassociated with negated and affirmative sentences containing hand-action verbs. Our results show reduced correlations forsentences where the verb is in the negated context, as compared to the affirmative one, within brain regions implicated inaction-semantic processing. The results lend support to the idea that negation involves reduced access to aspects of theaffirmative representation and pave the way for further testing alternate distributional-based semantic models of negationagainst human semantic processing in the brain.