Lexical olfaction recruits olfactory orbitofrontal cortex in metaphorical and literal contexts

Abstract

The investigation of specific lexical categories has substantially contributed to advancing our knowledge on how meaning is neurally represented. One sensory domain that has received particularly little attention is olfaction. This study aims to investigate the neural representation of lexical olfaction. In an fMRI experiment, participants read olfactory metaphors, their literal paraphrases, and literal olfactory sentences. Regions of interest were defined by a functional localizer run of odor processing. We observed activation in secondary olfactory areas during metaphorical and literal olfactory processing, thus extending previous findings to the novel source domain of olfaction. Previously reported enhanced activation in emotion-related areas due to metaphoricity could not be replicated. Finally, no primary olfactory cortex was found active during lexical olfaction processing. We suggest that this absence is due to olfactory hedonicity being crucial to understand the meaning of the current olfactory expressions. Consequently, the processing of olfactory hedonicity recruits secondary olfactory areas.

Grounding language in the neglected senses of touch, taste, and smell

“…The results reviewed here suggest that mental simulation related to the lower senses may be difficult, and less common than for the "higher" senses of vision and audition (e.g., Luzzi et al, 2007;Pecher et al, 2003;Speed & Majid, 2018a;van Dantzig et al, 2008). The evidence for overlap in brain activation for sensory language and perception is tenuous at best (e.g., Goldberg et al, 2006aGoldberg et al, , 2006bPomp et al, 2018), with most studies unable to rule out the engagement of strategic mental imagery in their tasks. While there is stronger evidence for the activation of secondary sensory regions for taste and smell language, such activation suggests a different level of specificity than activation of primary olfactory and gustatory regions.…”

“…The remaining 27 sentences explicitly refer to an odour source (e.g.,"garlic" in Es ist schlecht für die Karriere, wenn man immer nach Knoblauch riecht; "It's bad for a career, always smelling of garlic."). Pomp et al (2018) suggest the lack of piriform activation is due to the overall hedonicity of the olfactory sentences-which highlight a pleasant or unpleasant odour experience-and therefore activate secondary olfactory regions where odour valence is processed. However, since valence is thought to be the primary dimension by which odours are encoded (Khan et al, 2007;Yeshurun & Sobel, 2010;Zarzo, 2008;although see Olofsson et al, 2012;Olofsson, Bowman, & Gottfried, 2013), and odour language is also strongly encoded along this dimension (Levinson & Majid, 2014;Winter, 2016;Wnuk & Majid, 2014) piriform activation ought to be expected on a fully grounded perspective.…”

“…A subsequent brain imaging study supports Speed and Majid's (2018a) conclusion that olfactory language activates representations of odour valence, but not low-level perceptual representations. Pomp et al (2018) conducted an experiment in German where participants read sentences that were either olfactory metaphors (e.g., Er kann ihn absolut nicht riechen; "He cannot smell him at all" meaning "He cannot stand him"), literal paraphrases (e.g., Er kann ihn absolut nicht ausstehen; "He cannot stand him at all"), or literal olfactory sentences (e.g., Er riecht sehr unangenehm; "He smells very unpleasantly"). Both literal and metaphorical olfactory sentences activated secondary olfactory regions of the brain (the orbitofrontal cortex; OFC), possibly reflecting odour valence (Rolls, 2004), but neither activated the piriform cortex.…”

“…On the other hand, one could argue the discrepancy lies in the fact that the sentence stimuli of Pomp et al (2018) do not refer to specific odours but general odour experiences (e.g., "He smells very unpleasantly"); but this would apply to only 8 out of 35 sentences. The remaining 27 sentences explicitly refer to an odour source (e.g.,"garlic" in Es ist schlecht für die Karriere, wenn man immer nach Knoblauch riecht; "It's bad for a career, always smelling of garlic.").…”

“…On the whole, then, there is no convincing evidence to date that odour-related language activates low-levels of olfactory processing in the absence of deliberative odour imaging, supporting the idea that odour and language are weakly connected in the brain for comprehension, as well as production, as has long been claimed (Engen, 1987;Lorig, 1999;Olofsson & Gottfried, 2015)-at least for the languages studied to date. Instead odour-related language may activate regions that process valence (Pomp et al, 2018;Speed & Majid, 2018a) at the level of broad odour categories (i.e., odours that are similar such as cinnamon and nutmeg; Speed & Majid, 2018a).…”