Pablo Solana
Centro de Investigación Mente, Cerebro y Comportamiento (CIMCYC), Universidad de Granada, España
(cc) Pablo Solana.
How do we understand the meaning of language? One of the most popular hypotheses nowadays is that people simulate the content of language in order to understand its meaning. In a curious experiment published in 2008, Glenberg et al. showed that moving beans in a specific direction hampered the processing of sentences describing movements in that direction. These results support the idea that we engage in motor simulations during language comprehension.
The ability to communicate thoughts and ideas through language is one of the most fascinating yet least understood human qualities. How do we extract meaning from the words and phrases of language? According to theories of embodied cognition, the meaning of concepts is closely linked to our bodily experiences (Barsalou, 2008). Thus, when we need to understand what a word like “draw” means, we can mentally simulate the movements needed to pick up a pencil and start making traces on a piece of paper, which helps us to better understand the meaning of the verb it refers to.
One of the first pieces of evidence in favor of this idea was the discovery of the Action-Sentence Compatibility Effect (ACE) by Glenberg and Kaschak in 2002. These authors presented their participants with sentences describing movements in two directions: “towards your body” (e.g., You open a drawer) and “away from your body” (e.g., You close a drawer). These sentences were presented interspersed with nonsense sentences (e.g., You write a drawer) and participants had to decide whether the sentences made sense or not. To do so, they pressed two buttons on a keyboard. In one half of the experiment, the “Yes” key was far away from the participant and the “No” key was close, while in the other half, the position of the keys was reversed. At the beginning of each trial, participants had to place their hand at a point equidistant from both buttons. In this way, responding with the far key involved a movement away from the body, while pressing the near key required a movement towards the body.
The results showed that, when participants processed an “away sentence”, they were quicker to respond if the response key was far away than if it was close. Conversely, if the sentence involved a “towards” movement, the pattern was reversed (Figure 1, left). These results were interpreted as evidence consistent with the existence of motor simulations during language comprehension: participants would have simulated the directionality of the actions described in the sentences, which made them faster at performing movements congruent with the simulated direction. However, this pattern of facilitation can also be explained by disembodied postures; for example, by a simple priming effect: the sentences could have activated the word “forward” or “backward”, and just by virtue of those words being congruent or incongruent with the response direction, the facilitation effect would have been generated, but without involving any kind of simulation (Mahon and Caramazza, 2008).
With this limitation in mind, Glenberg and colleagues (2008) designed a new and ingenious experiment. Participants performed a task very similar to the one described above. However, before starting this task, they were asked to perform a more curious task. Participants were given two boxes, one with 600 beans and one empty, and asked to transfer, one by one, all the beans from the full box to the empty one (an activity that lasted about 20 min). Crucially, half of the participants had the full box close by and the empty one far away, while the other half had the empty one close by and the full one far away. That way, participants had to spend 20 min performing movements towards their body or away their body.
According to embodied theories, understanding a verb describing a motor action partially requires the same motor resources as performing that action (Barsalou, 2008). Therefore, limiting the availability of these resources should negatively affect language processing. Hence the logic of asking to move the beans for a long time and in a specific direction: reducing the motor resources associated with movements in that direction. If people really do simulate the actions described in language, then “fatiguing” movements in a particular direction should make it difficult to understand sentences involving that directionality. Crucially, this result is difficult to explain by priming, which would predict just the opposite, which would more clearly support the existence of motor simulations than previous studies using the ACE paradigm.
The results supported the researchers’ hypothesis, showing a pattern opposite to that expected from a priming account. Those participants who moved the beans towards their body were slower to respond to sentences describing actions towards their body than sentences denoting movements from their body outwards. Conversely, those participants who moved the beans away from their body were slower responding to sentences describing actions from their body outwards than sentences denoting movements towards them (Figure 1, right). Moreover, this effect was observed both in sentences describing concrete actions (i.e., referring to a physical movement; e.g., “You open/close the drawer”) and in sentences referring to abstract actions (i.e., referring to a metaphorical movement; e.g., “You delegate responsibility to Mary/Mary delegates responsibility to you”).
Figure 1. On the left, results from Glenberg and Kaschack (2002). On the right, results from Glenberg et al. (2008). The vertical axis represents the reaction time in the sentence comprehension task. The horizontal axis and the color of the bars represent the experimental conditions of the studies. Note: Values are approximated from the results plots of these studies.
In conclusion, the results of Glenberg et al. (2008) support the idea that, during language processing, people generate motor simulations of its content to understand its meaning. Moreover, these simulations can occur even when language refers to concepts of an abstract nature, not based on bodily experience. This study is therefore considered one of the most conclusive in favor of embodied theories of language processing. Against all odds, the humble beans were of great use in understanding human cognition.
However, we should also be cautious about these curious results. To date, no published study has conceptually or directly replicated these findings. Notably, a recent study conducted simultaneously in 18 laboratories worldwide, with more than 1000 participants, failed to replicate the ACE effect that underlies this line of research (Morey et al., 2022). In light of this, future studies should therefore test the replicability and veracity of the results of Glenberg and colleagues (2008).
References
Barsalou, L. W. (2008). Grounded cognition. Annual Review of Psychology, 59, 617-645.
Glenberg, A. M., & Kaschak, M. P. (2002). Grounding language in action. Psychonomic Bulletin & Review, 9, 558-565.
Glenberg, A. M., Sato, M., & Cattaneo, L. (2008). Use-induced motor plasticity affects the processing of abstract and concrete language. Current Biology, 18, R290-R291.
Mahon, B. Z., & Caramazza, A. (2008). A critical look at the embodied cognition hypothesis and a new proposal for grounding conceptual content. Journal of Physiology-Paris, 102, 59-70.
Morey, R. D., Kaschak, M. P., Díez-Álamo, A. M., Glenberg, A. M., Zwaan, R. A., Lakens, D., … & Ziv-Crispel, N. (2022). A pre-registered, multi-lab non-replication of the action-sentence compatibility effect (ACE). Psychonomic Bulletin & Review, 29, 613-626.
Acknowledgements: SEPEX Grants for the Dissemination of Research 2023-2024. Project ref. PID2022-142583NB-I00, funded by MICIU/AEI/10.13039/501100011033 and FEDER, UE
Manuscript received on October 19th, 2024.
Accepted on February 27th, 2025.
This is the English version of
Solana, P. (2025). Lo que un puñado de judías nos enseñó sobre el lenguaje. Ciencia Cognitiva, 19:2, 44-47.
