Short CommunicationMetaphorically feeling: Comprehending textural metaphors activates somatosensory cortex
Highlights
► Texture-selective somatosensory cortex is known to be in the parietal operculum. ► Processing texture metaphors activates texture-selective somatosensory cortex. ► Consistent with the theory that metaphor comprehension is perceptually grounded.
Introduction
Some accounts of cognition propose that knowledge is represented in abstract codes, distinct from the sensory modalities through which the knowledge was acquired, and that cognitive processes involve computations on these amodal representations (Fodor, 1975, Pylyshyn, 2007). Theories of grounded cognition reject this notion, proposing instead that knowledge is represented in modal systems derived from perception and that cognition depends on perceptual simulations (Barsalou, 2008).
Conceptual metaphor theory is one approach to grounded cognition which suggests that knowledge is structured by metaphorical mappings from sensory experience (Lakoff & Johnson, 2003). It is argued that this is true even for abstract concepts like time, for which metaphorical mappings can be made from experience of more concrete domains, such as space (Boroditsky, 2000, Casasanto and Boroditsky, 2008); for example, we speak of falling ‘behind’ schedule or looking ‘forward’ to an event. Conceptual metaphor theory is supported by a number of behavioral studies (e.g. Ackerman et al., 2010, Boot and Pecher, 2010, Gibbs and Matlock, 2008, Ouellet et al., 2010, Wilkowski et al., 2009. For example, the experiences of light/heavy, rough/smooth, or hard/soft objects influenced subsequent judgments of importance, difficulty of social interaction, or flexibility in negotiation, respectively (Ackerman et al., 2010); and anger primes led to overestimation of actual room temperature (Wilkowski et al., 2009). These studies suggest that metaphorical expressions such as ‘weighty matters’, ‘coarse language’, ‘unbending attitude’, and ‘hot-headed’ have a perceptual basis. By contrast, others suggest that conventional metaphors, such as the time/space mappings noted above, are merely overlearned idiomatic associations and that lexicalization of such metaphors results in separate stored meanings (Keysar and Bly, 1999, Keysar et al., 2000). On this account, there is no need to invoke associative mapping to understand that having a ‘rough’ day means having a ‘bad’ day, since the metaphor is in such common usage that the meaning of the word ‘rough’ is stored as both ‘abrasive’ and ‘unpleasant’.
These two approaches to metaphor comprehension make very different predictions about the neural basis of metaphor processing. If lexicalized metaphors do not require access to the corresponding concrete concept, then activity should be largely limited to classical language areas, whereas conceptual metaphor theory predicts activity in sensory areas involved in processing the domain from which the metaphor is derived. Previous studies showed increased activity in language areas for both conventional (Eviatar and Just, 2006, Lee and Dapretto, 2006) and novel (Rapp, Leube, Erb, Grodd, & Kircher, 2004) metaphors. However, since the metaphors in these studies were drawn from multiple domains, the locus of sensory activity might have varied across trials, accounting for the failure to observe consistent sensory processing.
More recent studies have begun to address whether there is a sensorimotor basis for metaphor comprehension by restricting the metaphors employed to a single domain (Boulenger et al., 2009, Chen et al., 2008, Desai et al., 2011, Raposo et al., 2009). These studies have concentrated on action verbs, particularly in relation to the idea of embodiment. However, the picture that emerges from this work is not entirely consistent. Raposo et al. (2009) found activity in motor and premotor cortex for literal, but not metaphorical, usages of arm- or leg-related verbs. Conversely, Boulenger et al. (2009) found somatotopic activation for arm- and leg-related verbs in both literal and metaphorical usages. Aziz-Zadeh and Damasio (2008), in a critique of studies reporting somatotopic recruitment of motor regions by action verbs, suggested that alternative organizing principles such as action goals, rather than somatotopy, may govern the simulations presumed to underlie motor cortical recruitment. Recently, Desai et al. (2011) reported activity for both metaphorical and literal action sentences in the left anterior inferior parietal lobule and the cerebellum: this activity was interpreted as motor-related, although these regions also have non-motor functions. Interestingly, activity in primary and supplementary motor cortices and in the superior temporal sulcus (implicated in biological motion perception) was inversely related to familiarity of both metaphorical and literal usages (Desai et al., 2011). Metaphorical use of motion verbs recruited a part of the middle temporal gyrus, close to the visual motion area MT (Chen et al., 2008) but, since this study did not include a functional localizer, the precise relationship between regions processing motion perception and motion metaphors remains unclear.
We reasoned that the segregated processing of object properties in sensory cortex offers alternative approaches to testing conceptual metaphor theory. In the present study, we restricted metaphors to the single domain of texture, which is commonly perceived both haptically and visually. We previously demonstrated texture-selective regions in somatosensory cortex of the parietal operculum for haptic stimuli and in early visual cortex for both haptic and visual stimuli (Sathian et al., 2011, Stilla and Sathian, 2008). Here, we used rapid event-related fMRI to compare processing of sentences containing familiar, textural metaphors against literal sentences with similar meanings and sentence structures. We present preliminary evidence for the hypothesis that processing textural metaphors activates texture-selective somatosensory areas defined on independent functional localizer scans.
Section snippets
Results
Mean response time (±SEM) for metaphorical sentences (.84 s ± .12 s) was longer than for literal sentences (.63 s ± .07 s), 2-tailed t6 = −3.25, p = .02). A contrast of the metaphorical and literal conditions showed that the textural metaphors activated somatosensory cortex in the parietal operculum, in bilateral OP1 and left OP3, which were previously shown to be texture-selective during haptic perception in a different group of subjects (Stilla & Sathian, 2008), and also cortex around the right inferior
Discussion
Our findings provide the first clear evidence for activity in functionally localized, domain-specific sensory cortical areas during processing of metaphors. Using metaphors drawn from the single domain of texture and contrasting these with literal sentences matched for meaning and sentence structure, we observed activation in somatosensory texture-selective areas, but did not find differential activation either in language areas or in visual or bisensory texture-selective areas. Although
Participants
Seven people took part in the study (2 males, 5 females; mean age 20 years, 8 months). Due to the nature of the task, we excluded participants for whom American English was a second/non-native language. All were right-handed based on the validated subset of the Edinburgh handedness inventory (Raczkowski, Kalat, & Nebes, 1974). All were people for whom we had localized visual, haptic and bisensory texture-selective areas in a separate fMRI study of texture perception (Gibson et al., 2008). At that
Acknowledgments
Support to KS from the National Eye Institute, National Science Foundation and the Veterans Administration is gratefully acknowledged. We thank Amy Anderson for assistance in data collection.
References (49)
- et al.
Embodied semantics for actions: Findings from functional brain imaging
Journal of Physiology–Paris
(2008) Metaphoric structuring: Understanding time through spatial metaphors
Cognition
(2000)- et al.
Time in the mind: Using space to think about time
Cognition
(2008) - et al.
Functional-anatomical organization of predicate metaphor processing
Brain & Language
(2008) - et al.
Brain correlates of discourse processing: An fMRI investigation of irony and conventional metaphor comprehension
Neuropsychologia
(2006) - et al.
Brain areas underlying visual mental imagery and visual perception: An fMRI study
Cognitive Brain Research
(2004) - et al.
What role does multisensory integration play in the visuotactile perception of texture?
International Journal of Physiology
(2003) - et al.
Swimming against the current: Do idioms reflect conceptual structure?
Journal of Pragmatics
(1999) - et al.
Conventional language: how metaphorical is it?
Journal of Memory and Language
(2000) - et al.
Art for reward’s sake: Visual art recruits the ventral striatum
NeuroImage
(2011)