“…The literature search was conducted in accordance with PRISMA guidelines (Moher et al, 2009). Relevant studies published after 2000 were identified in NeuroSynth , Google Scholar , APA PsycInfo , and PubMed databases using Boolean searches of the following keywords: “anaphora,” “anthropomorphism,” “comedy,” “discourse comprehension,” “figurative language,” “figure of speech,” “hyperbole,” “humor”, “idioms,” “indirect request,” “indirect speech,” “ironic,” “irony,” “jokes,” “lying,” “metaphor,” “metonymy,” “narrative,” “non-literal language,” “oxymoron,” “paradox,” “personification,” “platitude,” “pragmatics,” “prosody,” “proverbs,” “pun,” “sarcasm,” “sarcastic,” “saying,” “speech act,” “synecdoche,” “text coherence,” “text comprehension,” “understatement”, “fMRI,” “brain,” and “neuroimaging.” We additionally examined the reference lists of past neuroimaging meta-analyses on non-literal language processing to minimize the possibility of missing relevant studies (Ferstl et al, 2008; Bohrn et al, 2012; Rapp et al, 2012; Vartanian, 2012; Vrticka et al, 2013; Lisofsky et al, 2014; Yang, 2014; Yang & Shu, 2016; Reyes-Aguilar et al, 2018; Farkas et al, 2021). (Note that we had originally planned to use the NeuroSynth database of extracted peaks (Yarkoni et al, 2011); however, we discovered that NeuroSynth automatically extracts peaks from SPM-style tables in published papers, and such tables frequently include peaks for contrasts in both directions: e.g., in our case, non-literal>literal and literal>non-literal.…”