Last reviewed on October 31, 2026.
For much of its early history, cognitive science treated the mind as a kind of software running on the brain's hardware. Perception delivered inputs; central reasoning operated on internal symbols; outputs were sent to the body. The body, on this picture, was peripheral — a delivery mechanism for percepts and a transport for action.
What is embodied cognition? (definition)
Embodied cognition is the theory that thinking is not just computation inside the brain: it depends on the body, its sensory and motor systems, and its ongoing interaction with the environment. Everyday examples include gesturing to help yourself explain an idea, counting on your fingers, and using a notebook or phone as an extension of memory. Some of its best-known experiments have not replicated, and how far the idea reaches is still debated.
Embodied cognition is the view that this picture leaves out something essential. It claims that what brains do depends in important ways on the bodies they inhabit, the actions those bodies can perform and the environments those bodies move through. The claim sounds modest. In practice it has reorganised research programmes in perception, language, development and robotics, and it remains one of the live debates in contemporary cognitive science.
What embodied cognition actually claims
The label covers a spectrum, and not every "embodied" claim is a strong one. Wilson (2002) separated six different claims made under the label; for a first pass it helps to separate four versions, ordered from least to most controversial.
- Body as constraint. Cognition has to work with the body's sensors, effectors and energy budget. This is uncontroversial; even classical cognitive science acknowledges it.
- Body as resource. The body is part of how problems get solved — gestures help thinking, eye movements scaffold attention, written notes offload memory. The mind uses the body, not just commands it.
- Concepts grounded in sensorimotor experience. Abstract concepts inherit structure from concrete bodily experience. On this view, understanding "grasp" or "fall" partly involves simulating the sensorimotor traces of grasping or falling (Barsalou, 2008).
- Cognition extended into the world. Some of what counts as thinking happens outside the brain entirely — in tools, in social practices, in the structure of the environment. The "extended mind" position of Andy Clark and David Chalmers (1998) is the strongest version.
A reader can accept the first two without committing to the second two. Many researchers do exactly that.
Why this is more than a rebrand
The reason embodied cognition matters is not that it added a new module to the standard model. It is that, when taken seriously, it changes what counts as an explanation. Classical cognitive science explains a cognitive feat by specifying the internal representations and rules that produce it. Embodied accounts often dissolve the need for some of those representations by showing that the body and the environment are doing work that was previously attributed to the brain.
A clear example is walking. A conventional walking robot computes and controls its joint movements continuously. McGeer (1990) showed that a machine with no motors or controller at all can walk down a gentle slope, because the mechanics of its legs produce the gait. Collins and colleagues (2005) built powered robots on the same principle that walk efficiently with little control. Part of the work that seemed to need computation is done by the body. Whether this generalises to higher cognition is exactly what the embodied research programme is trying to find out.
Evidence that has been taken seriously
Several lines of work are commonly cited.
Sensorimotor activation during conceptual processing
Hauk, Johnsrude and Pulvermüller (2004) found that reading action words such as "kick," "lick" and "pick" produces activity in motor and premotor areas associated with the corresponding body parts. This is consistent with the view that conceptual understanding partially re-engages the systems used in actually performing actions. Critics point out that activation does not entail necessity (Mahon & Caramazza, 2008); whether motor areas are causally required for understanding action language remains open.
Gesture and mathematical thinking
Children who were told to gesture during a maths lesson learned more than children who were not, and correct gestures helped most (Goldin-Meadow, Cook & Mitchell, 2009). Gesturing while explaining maths also seems to free up working memory (Goldin-Meadow et al., 2001). Adults encouraged to gesture solved more mental rotation problems, and the benefit lasted after gesturing stopped (Chu & Kita, 2011). One reading is that gesture is part of the cognitive process, not just an outward symptom of it (Novack & Goldin-Meadow, 2015).
Infant motor development and concept learning
What infants understand about space changes as their motor abilities change. Adolph (2000) found that nine-month-olds avoided reaching over a risky gap when sitting, a posture they knew, but fell in when tested in the less familiar crawling posture, which suggests the learning is tied to the posture and not to general knowledge about drops. Research by Karen Adolph and others shows that motor milestones change what infants attend to, try to do and learn (Adolph & Hoch, 2019); Campos and colleagues (2000) link self-produced movement to a range of perceptual and cognitive changes. Cognitive development is harder to explain without taking the changing body into account.
Tools and the extended mind
Monkey and human studies suggest that tool use can change how the brain represents the space around the hand (Maravita & Iriki, 2004). The evidence is less firm than the popular idea that tools become part of the body: a re-analysis by Holmes (2012) did not support the claim that tool use actually extends peripersonal space, and found small effects and low statistical power. This is the empirical end of the extended-mind philosophical position: even if you do not buy that a notebook is part of your mind, you cannot easily deny that the tools you use shape what you can think.
Which classic embodiment studies have replicated?
Several of the best-known findings in this area come from small studies that later failed in larger, preregistered tests. The status of the main ones is below. A failed replication does not sink embodied cognition as a whole, but these examples should not be quoted as established facts.
| Study | Original claim | Current status |
|---|---|---|
| Pen in the mouth (Strack, Martin & Stepper, 1988) | Holding a pen in the teeth, which forces a smile, makes cartoons seem funnier. | Not replicated in its original form. Seventeen labs found a difference of 0.03 points on a 10-point scale, against 0.82 in the original (Wagenmakers et al., 2016). Noah, Schul and Mayo (2018) argued that the replications’ video cameras removed the effect. The wider idea of facial feedback survives as a small effect: a meta-analysis found it small and variable (Coles, Larsen & Lench, 2019), and a 3,878-person study across 19 countries found that mimicking or voluntarily making a smile can raise happiness, with less conclusive results for the pen task (Coles et al., 2022). |
| Warm coffee (Williams & Bargh, 2008) | Holding a warm cup makes you judge other people as warmer. | Not replicated. A multi-lab replication failed (Lynott et al., 2014), and two further attempts favoured no effect (Chabris et al., 2019). |
| Power posing (Carney, Cuddy & Yap, 2010) | Holding an expansive pose raises testosterone, lowers cortisol and increases risk taking. | Hormone and behaviour effects not replicated. A study of 200 people found an effect only on self-reported feelings of power (Ranehill et al., 2015). Simmons and Simonsohn (2017) judged the published evidence too weak to recommend the practice; Cuddy, Schultz and Fosse (2018) dispute this. Dana Carney, the first author, said in 2016 that she no longer believed the effects were real, while retaining confidence in the effect on feelings of power (NPR, 2016). |
| Action-sentence compatibility (Glenberg & Kaschak, 2002) | Responses are faster when your movement matches the direction described in a sentence. | Not replicated. In a preregistered study, none of 18 labs found a reliable effect, and the pooled estimate was close to zero (Morey et al., 2022). The original authors took part. |
| Motor activity for action words (Hauk et al., 2004) | Reading action words activates related motor areas. | The activation is widely reported; its meaning is disputed. It may not be needed for understanding (Mahon & Caramazza, 2008). |
| Gesture and learning (Goldin-Meadow et al., 2009; Chu & Kita, 2011) | Gesturing helps people think and learn. | Better supported. Separate research groups have found benefits, including in experiments that manipulated gesture directly (Novack & Goldin-Meadow, 2015). |
The strongest objections
Embodied cognition has serious critics, and the criticisms are worth understanding.
- The activation-is-not-explanation objection. Showing that motor cortex lights up when someone reads "kick" does not show that motor cortex is doing the conceptual work. Activation might be a downstream consequence of comprehension achieved elsewhere.
- The abstract-concept problem. If concepts are grounded in sensorimotor experience, what about concepts like "justice," "infinity" or "modal logic"? Embodied accounts can extend grounding via metaphor, but the further the concept is from concrete experience, the more strained the story becomes.
- The replication concern. As the table above shows, several prominent embodied effects, particularly those in which subtle bodily cues supposedly shape judgements, have replicated poorly. The related behavioural-priming literature has had the same trouble (Doyen et al., 2012). This has not refuted the broader programme but has tempered its more sweeping claims.
- The "where do you stop?" worry. If thinking includes the notebook and the calculator, why not also the internet, the social institution, the entire culture? Adams and Aizawa (2001) argue that the extended-mind story confuses what causes or supports cognition with cognition itself.
What this means for AI and design
Outside cognitive science proper, embodied ideas have had concrete consequences.
In artificial intelligence and robotics, the embodied perspective fuelled the rise of behaviour-based robotics, in which intelligent behaviour emerges from the interaction of simple sensorimotor loops with a structured environment, rather than from a central planner (Brooks, 1991). Reinforcement-learning agents that act in physical or simulated worlds also inherit some of the embodied programme's commitments, even when they are not framed in those terms. Conversely, large language models, trained largely on text and without bodies, are a useful test case. They show competence on many tasks while having no sensorimotor grounding at all, which has rekindled the debate about how much of human-style understanding requires embodiment.
In design and human factors, the embodied stance underwrites a class of insights that applications already touches on: that good interfaces work with the body's habits rather than against them, that physical and digital tools are continuous, and that the environment can carry cognitive load that designers might otherwise try to put inside the user's head.
Where embodied cognition fits in the field
It is most accurate to think of embodied cognition not as the opposite of classical cognitive science but as a corrective, and a productive one. Few researchers today defend the strict view that cognition is purely symbolic computation insulated from the body; few defend the strict view that there are no internal representations at all. The interesting work happens in between, asking which cognitive capacities are best explained by internal mechanisms, which by sensorimotor coupling, which by the structure of the environment and which by all three together.
The disciplines overview situates embodied cognition relative to the older paradigms; the history page explains how the field arrived at this question; and the glossary defines terms such as representation, embodiment and extended mind that recur in this debate.
Frequently Asked Questions
What is embodied cognition in simple terms?
Embodied cognition is the idea that we think with our bodies as well as our brains. How we perceive, understand concepts and solve problems is shaped by having a body that moves, senses and acts in the world.
What is an example of embodied cognition?
Common examples are children counting on their fingers, people gesturing while explaining a spatial idea, and athletes 'reading' a situation through movement. Offloading memory to notes or a phone is an example of the extended version of the idea.
Who developed the theory of embodied cognition?
Key contributors include Francisco Varela, Evan Thompson and Eleanor Rosch (The Embodied Mind, 1991), George Lakoff and Mark Johnson on metaphor, Lawrence Barsalou on grounded cognition, Rodney Brooks in robotics, and Andy Clark and David Chalmers on the extended mind.
What is the difference between embodied cognition and the extended mind?
Embodied cognition says cognition depends on the body. The extended mind thesis goes further, claiming that tools and the environment, such as a notebook, can be part of the thinking process.
What are the main criticisms of embodied cognition?
Critics argue that some famous embodiment effects have failed to replicate, that motor activation during language may accompany understanding rather than cause it, and that strong versions blur the line between cognition and its causes.
Have the classic embodied cognition studies replicated?
Several have not. The pen-in-mouth smile study failed in a 17-lab replication, the warm-coffee study failed in later attempts, the hormone and behaviour effects of power posing did not replicate, and an 18-lab test found no action-sentence compatibility effect. Facial feedback itself seems to exist as a small effect, and gesture research has held up better.
Sources
- Wilson, M. (2002). Six views of embodied cognition. Psychonomic Bulletin & Review, 9(4), 625–636. doi:10.3758/BF03196322
- Varela, F. J., Thompson, E., & Rosch, E. (1991). The Embodied Mind: Cognitive Science and Human Experience. MIT Press.
- Lakoff, G., & Johnson, M. (1980). Metaphors We Live By. University of Chicago Press.
- Barsalou, L. W. (2008). Grounded cognition. Annual Review of Psychology, 59(1), 617–645. doi:10.1146/annurev.psych.59.103006.093639
- Clark, A., & Chalmers, D. (1998). The extended mind. Analysis, 58(1), 7–19. doi:10.1093/analys/58.1.7
- Adams, F., & Aizawa, K. (2001). The bounds of cognition. Philosophical Psychology, 14(1), 43–64. doi:10.1080/09515080120033571
- Brooks, R. A. (1991). Intelligence without representation. Artificial Intelligence, 47(1–3), 139–159. doi:10.1016/0004-3702(91)90053-M
- McGeer, T. (1990). Passive dynamic walking. The International Journal of Robotics Research, 9(2), 62–82. doi:10.1177/027836499000900206
- Collins, S., Ruina, A., Tedrake, R., & Wisse, M. (2005). Efficient bipedal robots based on passive-dynamic walkers. Science, 307(5712), 1082–1085. doi:10.1126/science.1107799
- Hauk, O., Johnsrude, I., & Pulvermüller, F. (2004). Somatotopic representation of action words in human motor and premotor cortex. Neuron, 41(2), 301–307. doi:10.1016/S0896-6273(03)00838-9
- 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(1–3), 59–70. doi:10.1016/j.jphysparis.2008.03.004
- Goldin-Meadow, S., Nusbaum, H., Kelly, S. D., & Wagner, S. (2001). Explaining math: gesturing lightens the load. Psychological Science, 12(6), 516–522. doi:10.1111/1467-9280.00395
- Goldin-Meadow, S., Cook, S. W., & Mitchell, Z. A. (2009). Gesturing gives children new ideas about math. Psychological Science, 20(3), 267–272. doi:10.1111/j.1467-9280.2009.02297.x
- Chu, M., & Kita, S. (2011). The nature of gestures’ beneficial role in spatial problem solving. Journal of Experimental Psychology: General, 140(1), 102–116. doi:10.1037/a0021790
- Novack, M., & Goldin-Meadow, S. (2015). Learning from gesture: how our hands change our minds. Educational Psychology Review, 27(3), 405–412. doi:10.1007/s10648-015-9325-3
- Adolph, K. E. (2000). Specificity of learning: why infants fall over a veritable cliff. Psychological Science, 11(4), 290–295. doi:10.1111/1467-9280.00258
- Adolph, K. E., & Hoch, J. E. (2019). Motor development: embodied, embedded, enculturated, and enabling. Annual Review of Psychology, 70(1), 141–164. doi:10.1146/annurev-psych-010418-102836
- Campos, J. J., Anderson, D. I., Barbu-Roth, M. A., Hubbard, E. M., Hertenstein, M. J., & Witherington, D. (2000). Travel broadens the mind. Infancy, 1(2), 149–219. doi:10.1207/S15327078IN0102_1
- Maravita, A., & Iriki, A. (2004). Tools for the body (schema). Trends in Cognitive Sciences, 8(2), 79–86. doi:10.1016/j.tics.2003.12.008
- Holmes, N. P. (2012). Does tool use extend peripersonal space? A review and re-analysis. Experimental Brain Research, 218(2), 273–282. doi:10.1007/s00221-012-3042-7
- Strack, F., Martin, L. L., & Stepper, S. (1988). Inhibiting and facilitating conditions of the human smile. Journal of Personality and Social Psychology, 54(5), 768–777. doi:10.1037/0022-3514.54.5.768
- Wagenmakers, E.-J., et al. (2016). Registered Replication Report: Strack, Martin, & Stepper (1988). Perspectives on Psychological Science, 11(6), 917–928. doi:10.1177/1745691616674458
- Noah, T., Schul, Y., & Mayo, R. (2018). When both the original study and its failed replication are correct: feeling observed eliminates the facial-feedback effect. Journal of Personality and Social Psychology, 114(5), 657–664. doi:10.1037/pspa0000121
- Coles, N. A., Larsen, J. T., & Lench, H. C. (2019). A meta-analysis of the facial feedback literature. Psychological Bulletin, 145(6), 610–651. doi:10.1037/bul0000194
- Coles, N. A., et al. (2022). A multi-lab test of the facial feedback hypothesis by the Many Smiles Collaboration. Nature Human Behaviour, 6(12), 1731–1742. doi:10.1038/s41562-022-01458-9
- Williams, L. E., & Bargh, J. A. (2008). Experiencing physical warmth promotes interpersonal warmth. Science, 322(5901), 606–607. doi:10.1126/science.1162548
- Lynott, D., Corker, K. S., Wortman, J., Connell, L., Donnellan, M. B., Lucas, R. E., & O’Brien, K. (2014). Replication of “Experiencing physical warmth promotes interpersonal warmth” by Williams and Bargh (2008). Social Psychology, 45(3), 216–222. doi:10.1027/1864-9335/a000187
- Chabris, C. F., Heck, P. R., Mandart, J., Benjamin, D. J., & Simons, D. J. (2019). No evidence that experiencing physical warmth promotes interpersonal warmth. Social Psychology, 50(2), 127–132. doi:10.1027/1864-9335/a000361
- Carney, D. R., Cuddy, A. J. C., & Yap, A. J. (2010). Power posing. Psychological Science, 21(10), 1363–1368. doi:10.1177/0956797610383437
- Ranehill, E., Dreber, A., Johannesson, M., Leiberg, S., Sul, S., & Weber, R. A. (2015). Assessing the robustness of power posing. Psychological Science, 26(5), 653–656. doi:10.1177/0956797614553946
- Simmons, J. P., & Simonsohn, U. (2017). Power posing: P-curving the evidence. Psychological Science, 28(5), 687–693. doi:10.1177/0956797616658563
- Cuddy, A. J. C., Schultz, S. J., & Fosse, N. E. (2018). P-curving a more comprehensive body of research on postural feedback reveals clear evidential value for power-posing effects. Psychological Science, 29(4), 656–666. doi:10.1177/0956797617746749
- NPR (2016, October 1). Power poses co-author: ‘I do not believe the effects are real’. npr.org
- Glenberg, A. M., & Kaschak, M. P. (2002). Grounding language in action. Psychonomic Bulletin & Review, 9(3), 558–565. doi:10.3758/BF03196313
- Morey, R. D., et al. (2022). A pre-registered, multi-lab non-replication of the action-sentence compatibility effect (ACE). Psychonomic Bulletin & Review, 29(2), 613–626. doi:10.3758/s13423-021-01927-8
- Doyen, S., Klein, O., Pichon, C.-L., & Cleeremans, A. (2012). Behavioral priming: it’s all in the mind, but whose mind? PLoS ONE, 7(1), e29081. doi:10.1371/journal.pone.0029081