Human language is a strange phenomenon. Somehow, we’re able to convey complex ideas through a fuzzy communicative channel. Even disregarding the remarkable machinery involved in transforming sound waves into neural signals, how does meaning emerge from those signals? And how do we talk about abstract concepts like “Justice”, “Truth”, or even “Concepts” themselves?
In recent years, cognitive scientists have found evidence that people understand language by mentally simulating the events it describes [1, 2, 3, 4, 5]. For example, when we hear the sentence the man grasped the baseball bat, the same parts of our brain activated when we actually grasp a baseball bat are activated. This is called the embodied simulation theory, and it predicts that for sentences describing concrete events – e.g. ones involving physical action or perception – we activate similar parts of our brain as we would if we observed those events or participated them ourselves.
So if we understand “concrete” sentences by simulating them, how do we understand “abstract” sentences?
The theory of embodied simulation connects quite naturally with Conceptual Metaphor Theory, which argues that we understand many higher-level concepts through the use of embodied metaphors . This manifests in everyday language in surprising ways. For example, it’s almost impossible to talk about TIME without talking about SPACE and MOTION: the years are moving by so fast, we’re fast approaching Christmas, or even you have such a bright future ahead of you. Indeed, experimental evidence suggests that we do think about TIME in terms of SPACE . Metaphors have been shown to affect reasoning as well; participants in one study made substantially different policy recommendations about reducing crime in a city when crime was conceptualized as a BEAST or VIRUS, respectively .
Together, these theories predict that people understand metaphors about higher-level concepts by simulating the more concrete actions they describe. For example, the public grasped the idea will involve activating parts of our brain responsible for grasping things, just as the man grasped the baseball bat does. And in fact, this has been demonstrated via neuro-imaging studies [9, 10], though another study found no activation for idiomatic expressions like “he bit off more than he can chew” .
So do people actually simulate metaphors about embodied actions? And if they do, to what extent? And if they don’t, how do they understand abstract language?
These are precisely the questions the researchers addressed in The neural career of sensory-motor metaphor .
Neural Career of Metaphor
To address these questions, the researchers recorded brain activity of 22 participants in an fMRI scanner while they read sentences in one of the following conditions:
- Literal: the daughter grasped the flower
- Metaphorical: the public grasped the idea
- Abstract: the public understood the idea
With a separate participant pool, each stimulus was assessed for its familiarity – e.g., how familiar the sentence seemed to participants.
Ultimately, the researchers found activation in regions responsible for motor actions in both the Literal and Metaphorical conditions, but not the Abstract condition. This makes sense, because both the literal and metaphorical conditions included language that referenced physical actions – and it’s consistent with the studies referenced above.
What’s more interesting is the finding that the degree of activation in these conditions – e.g., how much more activation was seen in comparison to the Abstract condition – was inversely correlated with a sentence’s “Familiarity rating”. That is, more familiar sentences seemed to require less simulation. The relationship between familiarity and degree of activation can be seen in the (admittedly somewhat confusing) figure below. The basic lesson to draw from the figure is that there’s more activation in motor regions during comprehension of literal and metaphorical statements than the abstract statements – and furthermore, that activation in certain of these regions are negatively correlated with a sentence’s familiarity.
The authors conclude that metaphor comprehension (and literal sentences generally) involves some degree of understanding. Moreover, they argue that over time, a metaphor becomes more cognitively “entrenched”, and we don’t need to simulate it as much. They refer to this as a ‘gradual abstraction of embodiment’. This seems to me to be a reasonable stance, though fMRI data is messy and correlational, so it’s sometimes unwise to draw significant inferences in the absence of more causal data.
Still, the authors’ argument seems reasonable enough from an intuitive point of view. We shouldn’t need to simulate sentences as much if we’re more familiar with those types of sentence. Personally, I think of this as a kind of caching mechanism; once we become familiar with the meaning of a certain grouping of words, we don’t need to apply the same simulation function to understand the meaning each time we hear those words together. This is why more familiar sentences are easier to read and understand – they take less effort. In other words, our brain has to do less work.
There are, however, many remaining questions, all of which merit considerable research attention (in my opinion). I describe a few below:
First of all, assuming there is this kind of caching mechanism, how exactly does this work in the brain? That is, how is the correspondence between input and output represented?
Second, since purely “abstract” sentences seem to involve less simulation, what mechanism is responsible for their comprehension? Has the meaning already been “cached”, perhaps through previous simulations? Another way of putting this is: is there a way to understand “abstract” concepts without the use of embodied metaphors?
Third, what factors define the “need” to simulate? It seems to me that we must simulate a sentence in the absence of a well-defined cache of that sentence, but given the generally fuzzy nature of the brain, there’s probably some fuzziness around the factors influencing the degree of simulation. Perhaps more interestingly, what can this tell us about the roots of language?
Finally, how does the degree of simulation of a metaphor relate to how much the metaphor affects our reasoning? I mentioned above that metaphors can influence decisions we make (e.g. about crime and law enforcement policies); so if a metaphor is more familiar to us, does that mean it affects our reasoning more or less?
So if you’re interested in language and the brain, these are all questions worth considering. And if you’re not, I’m surprised you made it to the end of this post.
- Borreggine, K. L., & Kaschak, M. P. (2006). The Action – Sentence Compatibility Effect : It’s All in the Timing. Cognitive Science, 30(August 2009), 1097–1112. http://doi.org/10.1207/s15516709cog0000
- Glenberg, A. M., & Kaschak, M. P. (2002). Grounding language in action. Psychonomic Bulletin & Review, 9, 558–565
- Kaschak, M. P., Madden, C. J., Therriault, D. J.,Yaxley, R. H.,Aveyard, M., Blanchard, A. A., et al. (2005). Percep- tion of motion affects language processing. Cognition, 94, B79–B89
- Barsalou, L. W. (1999). Perceptual symbol systems. Behavioral and Brain Sciences, 22(4), 577-609–60. http://doi.org/10.1017/S0140525X99252144
- Boulenger, V., Mechtouff, L., Thobois, S., Broussolle, E., Jeannerod, M., Nazir, T.A., (2008). Word processing in Parkinson’s Disease is impaired for action verbs but not for concrete nouns. Neuropsychologia 46, 743–756
- Lakoff, G., & Johnson, M. (1980). Conceptual metaphor in everyday language. The Journal of Philosophy. http://doi.org/10.2307/2025464
- Boroditsky, L. (2000). Metaphoric structuring: Understanding time through spatial metaphors. Cognition, 75(1), 1–28. http://doi.org/10.1016/S0010-0277(99)00073-6
- Thibodeau, Paul; Boroditsky, L. (2011). Metaphors We Think With: The Role of Metaphor in Reasoning. PLoS ONE (2011), 89(3), 493–510. http://doi.org/10.1371/journal.pone.0016782
- Raposo A, Moss HE, Stamatakis EA, Tyler LK. (2009) Modulation of motor and premotor cortices by actions, action words and action sentences. Neuropsychologia. 47:388–396. [PubMed: 18930749]
- Boulenger V, Hauk O, Pulvermuller F. (2009) Grasping Ideas with the Motor System: Semantic Somatotopy in Idiom Comprehension. Cereb Cortex. 19:1905–1914. [PubMed: 19068489]
- Aziz-Zadeh L, Damasio A. (2008) Embodied semantics for actions: Findings from functional brain imaging. Journal of Physiology-Paris. 102:35–39.
- Desai, R. H., Binder, J. R., Conant, L. L., Mano, Q. R., & Seidenberg, M. S. (2011). The neural career of sensory-motor metaphors. Journal of Cognitive Neuroscience, 23(9), 2376–2386. http://doi.org/10.1162/jocn.2010.21596
 The reason this is interesting is that we think we have a good idea of how input (language) is mapped to output (meaning) when it involves simulation. Language input results in simulation, which essentially is the meaning of the language. (This is kind of a philosophical point, but given the fact that we’re all embodied beings moving through the world, seeing things and touching things and so on, our most basic experiences are impossible to dissociate from their embodiment, and so whatever ‘meaning’ is, it surely must involve some sort of relationship with these embodied experiences. In my opinion.)