Advanced Review

Embodiment as a unifying
perspective for psychology
Arthur M. Glenberg1,2

A basic claim of the embodiment framework is that all psychological processes are
influenced by body morphology, sensory systems, motor systems, and emotions.
As such, the framework holds the promise of providing a unifying perspective for
psychological research. This article begins with a sketch of several arguments, from
evolution to philosophy, as to why the embodiment framework is a good bet. These
arguments are followed by a review of approaches to embodiment, including those
from cognitive linguistics, perceptual symbol theory, and action-based theories.
Finally, examples are provided for how a unifying perspective might work for
cognition (including language and memory), cognitive and social development,
social psychology, neuroscience, clinical psychology, and psychology applied to
education. 2010 John Wiley & Sons, Ltd. WIREs Cogn Sci 2010 1 586596

W hy is progress slow in psychology? Perhaps it

is because there is so little agreement among
the content areas (e.g., cognitive psychology, develop-
body, including body morphology, sensory systems,
and motor systems. How can this starting point
provide a suitable beginning for the multiple concerns
mental psychology, social, and so on), or perhaps, as of psychologists? How can the body help when many
Mischel1 speculates, the drive for individual recogni- of these concerns seem to be unrelated to the body,
tion and theory development precludes a cumulative concerns such as abstract cognition, language and
advance. A third, but related, possibility is that areas communication, social processes, and psychological
do not talk the same language; they do not have the dysfunction, to name but a few?
same organizing principles and metaphors, and hence First, it is almost a certainty that psychological
it is difficult to ascertain commonalities in approach, processes or their underlying substrates (e.g., the
data, and theory. The embodied approach to psychol- ability to engage in high-level cognition), evolved.
ogy offers the possibility of unification of language, And, in the main, evolution is driven by the dual
problems, and theory. In this article, I begin with some imperatives of survival and reproduction. These
reasons as to why embodiment might be a unifying imperatives require direct interaction with the physical
force, followed by a review of several approaches to and social world, and that interaction is only through
embodied psychology. Finally, the main section of the the body. In fact, the biologist, Rudolfo Llinas2
article consists of a selective review of research show- asserts that A nervous system is only necessary
ing how embodiment provides coherence to research for multicellular creatures. . .that can orchestrate and
in psychology. express active movement. . . (page 15). That is, brains
are for guiding interaction with the world, and that
interaction is mediated by the body.
HOW CAN THE BODY UNIFY? Along the same lines, consider the necessity of
Although there are different versions of embodiment the coevolution of body and cognition. When faced
theory, for the most part, they take as a starting point with a predator, if a mole were to attempt to fly away,
that psychological processes are influenced by the that mole would not survive to contribute to the gene
pool. Similarly, if a bird attempted to dive into a hole
Correspondence to: [email protected] in the ground, it would not contribute to the gene pool.
1 Departmentof Psychology, Arizona State University, Phoenix, In general, how we analyze and react to a situation
AZ, 85287, USA must take into account bodily abilities. Given that
2 Departmentof Psychology, University of Wisconsin, Madison, a nervous system is only necessary for action, and
WI, 53706, USA given that action requires a body and consideration
DOI: 10.1002/wcs.55 of that bodys abilities, it is a good bet that many

586 2010 Jo h n Wiley & So n s, L td. Vo lu me 1, Ju ly /Au gu s t 2010

 WIREs Cognitive Science Embodiment as a unifying perspective

psychological processes have their roots (if not their meaning of the first word in the sign, let alone the
trunk, limbs, and leaves) in the need for action meaning of the sign itself. In short, symbols must be
Second, consider an argument from the philoso- grounded, that is, related to something other than
pher Merleau-Ponty (see Ref 3) concerning similarity. additional symbols. The sensory, action, and emotion
The argument is based on the idea that similarity is systems of our bodies provide that grounding: words,
related to the number of features in common for the phrases, and mathematical and logical symbols all
objects being judged: the more features in common, become meaningful through how we perceive and
the greater the similarity. But, if one is free to pick interact with the objects and situations those symbols
features, then any two objects or events can be made denote.
as similar as one wishes, that is, the two can have
an infinite number of features in common. For exam-
ple, a lion and my computer are similar in that they VARIETIES OF EMBODIMENT
both exist on earth, they do not exist on the moon, All approaches to embodiment agree that behavior is
they do not exist on Mars. . .they exist at a time point produced by more than a disembodied Cartesian mind
after 1,000,000 BCE, after 999,999.99 BCE, after manipulating symbols according to rules. In other
999,999.98 BCE, and so on. One might object, how- words, embodiment is in strong contrast to cognitive
ever, that these are arbitrary and silly features. But, psychology as developed in the 1960s, 1970s,
why are the features arbitrary and silly? It is because and 1980s.6,7 Various approaches to embodiment
they have little to do with human interactions with emphasize, to greater and lesser degrees, contributions
objects. Those interactions are guided by features such of evolution to the shaping of cognition, the
as perceived color, and how objects can be gripped, coevolution of body and behavior, culture, the
or eaten, or used. Note, however, that the humanly specific environment in which the organism is acting,
relevant features depend on having human bodies that emotional states, details of perceptual systems, action
perceive particular wavelengths and human bodies systems, and neural systems.
that effect particular actions (e.g., gripping and eat- Lakoff8 and Gallese and Lakoff9 provide one
ing). Thus, similarity, a basic and central component example of embodied meaning. Part of Lakoffs
of cognition and social interaction, depends on the project is to understand why so much language is
particulars of the body. metaphorical. For example, we talk about theories as
Finally, it is clear that humans are symbolic if they were buildings with a structure, a foundation,
creatures: our lives are dominated by the symbols of supporting members, and so on; and we talk about
language, and many societies spend enormous sums relationships as if they were journeys with beginnings,
educating their children in how to make meaning middles, ends, rocky parts, as well as smooth parts;
out of linguistic and mathematical symbols. But and we talk about mood and emotion as having
how do symbols work? At one time, psychological a spatial dimension as when a happy person is
theory asserted that symbols become meaningful by described as up or flying high and a sad person is
participating in a network of other symbols, such as down in the dumps. On Lakoffs analysis, the use of
semantic networks or high-dimensional spaces. It is metaphor is much more than just a way of talking;
now relatively well-accepted that this story wont do.4 instead metaphors reveal the way people represent
Consider, for example, Harnads5 symbol merry-go- and think about abstract concepts, and importantly,
round argument. Imagine that you just landed in a those representations result from literal interactions
foreign country, and you dont speak the language. of the body with the world. When people are sad,
At your disposal, you have a dictionary for that they literally slump, sit, or lie down, whereas when
language. As you enter the airport terminal, you someone is joyous, they literally carry themselves erect
see a sign written in the foreign language and you and may literally jump for joy. Thus, the way we think
decide to figure out its meaning. You look up the first about emotions may reflect the literal interactions of
word in the dictionary only to find that it is defined our bodies with the environment.
using other words in the language that you do not Another example from Lakoffs work is his
understand. Undaunted, you look up the first word in analysis of logical expressions such as either a or b,
the definition, and it too is defined only in terms of but not both. Much of our early experiences involve
words of the language you do not understand. One interactions with containers, such as cups, bottles,
can imagine looking up word after word after word, boxes, and even our own bodies. Part of the consistent
that is, tracing the relations in the semantic network experiences with containers involves putting things
defined by the dictionary. And yet, no matter how in and taking things out. According to Lakoff, the
many words you look up, you will never uncover the fact that these experiences have a consistent structure

Vo lu me 1, Ju ly /Au gu s t 2010 2010 Jo h n Wiley & So n s, L td. 587

 Advanced Review wires.wiley.com/cogsci

(e.g., something is in the container or out of the CRANBERRIES are tart), and that is just what was
container, but not both) results in an image schema found.
for containers, and the structure of the schema reflects A third approach has emphasized the contribu-
our bodily interactions with containers. Finally, we tion of action to cognition and meaning. For example,
learn to metaphorically extend the basic, embodied Glenberg and Kaschak16 tested the hypothesis that
schemas toward more abstract concepts that have the sentence understanding involves a simulation not
same structure. In this way, we come to understand just of perceptual qualities but also of action. They
abstract logical expressions such as either a or b, but asked participants to judge the sensibility of sentences
not both as container-like, in that something can be describing transfer away from the participant (e.g.,
in or out of the container, but not both. You give Art the pencil) or toward the participant
The notion that abstract cognition metaphor- (e.g., Art gives you the pencil) in contrast to a non-
ically hinges on bodily activity has been further sense sentence (e.g., You give the pencil Art). To
developed and tested by Gibbs10 and Boroditsky indicate that a sentence was sensible, half of the par-
and Ramscar.11 Lindeman and Abramson12 apply a ticipants moved the hand to a response button away
Lakoffian analysis in clinical psychology, as described from the body and half moved the hand toward the
below. body. The major result was that participants were
A second approach to embodied cognition is faster to judge a sentence as sensible when the implied
Barsalous13 notion of a perceptual symbol system. direction of the sentence (toward or away) matched
In contrast to the standard cognitive science notion the literal direction of the response.
of symbols as abstract, point-like entities, perceptual Glenberg et al.17 had participants judge sensibil-
symbols are analogical in the sense that they are ity of transfer sentences by simple key presses using
composed of components of neural activity arising two fingers of the left hand without hand movements.
from the perception of the symbols referent. For Transcranial magnetic stimulation was used to mea-
example, a perceptual symbol of a car might consist sure any effects of sentence processing on right hand
of neural activity in the visual system that arises muscles. Sentences that described transfer produced
from seeing a car, neural activity in the auditory more muscle activation than sentences that did not
system corresponding to the sound of a car, and so (e.g., You and Art look at the pencil), as if the neuro-
on. The perceptual symbol is not an exact copy of muscular system used in literal grasping and transfer
the neural activity arising from perception; instead it was also used in simulating transfer. Importantly,
consists of the neural activity underlying the attended the same results were found for sentences describing
components of the experience. transfer of information (e.g., You delegate the respon-
An important idea associated with perceptual sibilities to Anna). That is, even the understanding
symbol systems is that they are used in simulation of abstract sentences seems to involve a sensorimotor
or reenactment. It is the simulation, constrained by simulation.
particular contextual features (e.g., the location of Zwaan and Taylor18 developed another pro-
the car, its speed, etc.), that corresponds to particular cedure for studying the role of action in language
exemplars of a category. comprehension that allows a finer-grain analysis. In
Barsalous approach has generated a tremendous their experiments, participants turned a knob either
amount of research, most of which is consistent with clockwise or counterclockwise to advance through
the theory. One of the most striking findings was a text. Some sentences described an action that is
reported by Pecher et al.14 They noted that there typically clockwise (increase the volume on a radio)
are switching costs when moving attention from or counterclockwise (decrease the volume). Zwaan
one modality to another. For example, in detecting and Taylor were able to demonstrate that the major
whether a stimulus is on the left or the right, if the source of interference between literal movement and
previous stimulus and the current stimulus are both implied movement was at the verb, although the dura-
visual, participants are faster than if the previous tion of interference could be extended if the sentence
stimulus is auditory and the current stimulus visual.15 continued to focus on the action.
If perceptual symbols are constructed from activity The experiments by Glenberg et al. and Zwaan
in different sensory systems, then there should be and Taylor are consistent with results from neural
switching costs when making conceptual judgments imaging. For example, Hauk et al.19 demonstrated
as well. For example, in verifying properties such as a greater activation of motor cortex controlling the
BLENDER is loud, people should be faster when the hand while listening to verbs such as pick, and greater
previous trial was in the same modality (e.g., LEAVES activation of motor cortex controlling the leg while
are rustling) compared to a different modality (e.g., listening to kick.

588 2010 Jo h n Wiley & So n s, L td. Vo lu me 1, Ju ly /Au gu s t 2010

 WIREs Cognitive Science Embodiment as a unifying perspective

In the following sections, I present examples of Another example, with older infants, was
how principles of embodiment, such as those discussed reported by Campos et al.21 When an infant is
above, can be applied to development, language being carried, there is little correlation between self
and memory, emotion and social psychology, theory movement (e.g., of the arms and legs) and changes
of mind, psychological disorders, and educational in the visual information, such as optical flow.
psychology. Each section begins with a claim about However, when the infant learns to locomote, there
the bodys contribution to the psychological processes is a strong correlation between action and optical
and then a review of some of the relevant research. flow that is used in guiding direction of travel and
maintaining balance. It is also the case that infants
with little self-locomotion experience show little fear
in crossing the visual cliff. That is, these infants can be
UNIFYING PSYCHOLOGY: COGNITIVE
induced to cross a glass-covered opening on a raised
AND SOCIAL DEVELOPMENT crawling surface. Infants with extensive experience
The embodiment claim is that cognitive and social will not cross the cliff. Campos reasoned that for self-
development is driven by physical development and locomoting infants, the cliff creates a discontinuity in
associated changes in action. That is, as infants the correlation between motion and optic flow, as if
learn control over different types of movement, the world is operating under different rules, which in
the infants literally generate different information turn generates fear. To test this claim, prelocomoting
structures for themselves that can change both infants were given experience sitting in a wheeled
cognition and emotion. One example is provided device that the infants could move by kicking against
by an investigation of patterns of attention.20 In the floor. Once these infants had experience using this
that research, 3-month-old infants observed a hand device (and in building the correlation between loco-
reaching for one of two objects in a particular location. motion and optic flow), they too displayed fear of the
This procedure continued until the infant stopped visual cliff. Thus, learning a new form of bodily activ-
looking (habituated). Then, infants were shown two ity, self-locomotion, changes the way the infant sees
new events. In the new goal event, the hand reached for the world and affects emotional and social responses.
the other object (a new goal for the reach), but at the Smith22 examined the relation between action
old location. In the new path event, the hand reached and categorization. In her experiment, 3-year-old
for the original object, but at a new location. The toddlers were given a slightly asymmetric, elongated
question is what were the infants attending during object and told it was a wug. Half of the toddlers
habituation: the goal or the path? If infants were were induced to play with the object in a symmetric
attending to the path (i.e., the physical movement), fashion by using both hands to hold the wug and
then the new path should be a change and produce rotate it back and forth using symmetric wrist turns.
dishabituation and longer looking times. In contrast, if The other toddlers were induced to play with the
the infants were attending the goal, then the new goal wug using an asymmetric action: holding the wug in
event should be particularly interesting and produce one hand and waving it. After playing, the toddlers
longer looking times. were asked if other objects were wugs. Some of these
The key to the Somerville et al. experiment was other objects were more symmetric than the original
what infants did before the habituation trials. Half of and some were less symmetric. The infants who had
the infants were given experience wearing and using played with the original using a symmetric action
sticky mittens, that is, mittens with Velcro that could were more likely to call the symmetric new objects
be swiped at toys with the matching Velcro. Thus, the wugs, whereas the infants who had played with the
mittens allowed infants to capture and examine toys original in an asymmetric fashion were more likely to
in ways that were not yet in the infants repertoire. call the asymmetric new objects wugs. Thus, the way
Amazingly, infants who had experience with the the toddler physically manipulates the object affects
mittens showed longer looking times to the new goal similarity judgments and categorization.
event compared to the new path event. Infants without
the experience showed no preference. That is, being
able to manipulate and examine objects changed the UNIFYING PSYCHOLOGY: LANGUAGE
nature of the infants attention from movements to
goals. This recognition of the goals underlying actions The embodiment claim for language is that sentences
is basic not only to cognitive development but also are understood by simulating sentence content using
to social development in that it allows the infant to neural systems ordinarily used for perception, action,
understand what others are doing. and emotion. Research described above demonstrates

Vo lu me 1, Ju ly /Au gu s t 2010 2010 Jo h n Wiley & So n s, L td. 589

 Advanced Review wires.wiley.com/cogsci

the connection between language comprehension and Havas et al.28 (see also Ref 29) adduced evidence
action (for more, see Ref 23). that the emotional system is involved in comprehend-
Kaschak et al.24 were among the first to demon- ing sentences describing emotional events. Participants
strate a role for the use of perceptual information were told that the experiment was investigating the
during sentence comprehension. In those experiments, effect of blocking the articulators on reading. Then,
participants listened to sentences describing motion participants either held a pen in the mouth using only
in a particular direction (e.g., The car approached the teeth, which forces a smile and brightens affect,30
you) and verified their sensibility. At the same time, or they held the pen using only the lips, which prevents
participants were looking at a visual stimulus that smiling and induces a frown. Havas et al. found that
appeared to be moving in the same direction as that participants were faster to comprehend sentences
implied in the sentence or the opposite direction. describing events likely to make one happy when they
The important finding was an interaction between were smiling, and that the participants were faster to
the direction of motion implied in the sentence and comprehend sentences describing events likely to make
the direction of motion in the visual stimulus in one sad when they were frowning. Apparently, having
determining the time taken to judge the sensibility the body in a matching emotional state facilitates
of the sentence. Apparently, understanding sentences comprehension of sentences describing emotionally
describing motion requires the same neural systems congruent events, just as if part of understanding the
that are used in perceiving motion. sentence depends on simulating the emotional state.
Vigliocco et al.25,26 have demonstrated striking
contributions of perceptual systems to language
understanding. Namely, hearing verbs implying visual UNIFYING PSYCHOLOGY: MEMORY
motion (e.g., rise) affects detection of literal motion, The embodiment claim regarding memory is that
and a motion detection task can interfere with memory reflects modality-specific and effector-specific
judgments of the lexical status of motion words. interactions with the world. As reviewed previously,
An embodied approach to language also Pecher et al.31 demonstrated just this in regard to
addresses aspects of language that have long been conceptual information in memory. The two studies
characterized as abstract (e.g., syntax and emotion) reviewed below demonstrate the claim for episodic
and hence beyond an embodied analysis. For exam- memory, that is, memory for particular episodes.
ple, Chambers et al.27 studied the relation between Brunel et al.32 demonstrated how particulars
tools, bodily capability, and syntactic analysis. Some of perceptual experience become incorporated into
types of clauses are ambiguous. For example, Move memories. In the first phase of the experiment,
the whistle on the table. . . might mean that a whistle participants were presented with 80 trials consisting
should be moved onto a table, or that the whistle that of 40 presentations of a square intermixed with 40
is already on the table should be moved elsewhere. presentations of a circle. One of these visual stimuli,
This latter interpretation is more common when there the critical stimulus, was associated with 500 ms of
are two whistles in the scene so that on the table white noise on each of the 40 trials. In the next phase
serves as a reduced relative clause that identifies a of the experiment, the two visual stimuli were used
particular whistle, namely, the one on the table. In as warning signals for a tone perception task. When
the experiment, participants literally followed instruc- the critical stimulus that had been associated with
tions to do things such as move whistles. In one the white noise was the warning signal, performance
condition, the hand was used to follow the instruc- on tone perception was reduced. Furthermore, this
tions, and in another condition, participants used a reduction only occurred when the tone was presented
hook. In addition, one whistle had a lanyard so that within 500 ms of the warning stimulus. Apparently,
it could be picked up using the hook. Here is the crit- the critical stimulus invokes a memorial representation
ical result. When participants were using their hands of the white noise using the same neural systems as
so that either of the two whistles could be moved, used in perception. Consequently, when the tone is
they interpreted on the table as a relative clause that presented during the duration of the memory (about
differentiated the two whistles. However, when par- 500 ms), performance is lowered because the neural
ticipants were holding the hook, they interpreted on system is already in use.
the table as the location to which the whistle with Yang et al.33 demonstrated how effector-specific
the lanyard (the only whistle that they could move) action information is used in memory. In their
should be moved. That is, the grammatical analysis of experiment, participants first memorized pairs of
the sentence was affected by the bodily capability to letters. Some pairs consisted of letters that could
move the whistles. be typed with the same finger, whereas the other

590 2010 Jo h n Wiley & So n s, L td. Vo lu me 1, Ju ly /Au gu s t 2010

 WIREs Cognitive Science Embodiment as a unifying perspective

pairs could be typed with different fingers on different on social cognition are provided by Mussweiller,35
hands (note, however, that there was no typing Jostmann et al.,36 and Semin and Smith.37
done in the experiment). When typing two letters Turning to emotion, consider first the mere expo-
with the same finger, one is less fluent than when sure effect:38 When novel stimuli are repeated, people
typing two letters with different fingers on different tend to rate the repeated stimuli as more likeable
hands, probably because the motor programs must than nonrepeated stimuli.a One well-supported expla-
be planned and executed serially in the former case. nation of the mere exposure effect is that repetition
In the second phase of the experiment, participants leads to more fluent processing, and that fluency leads
were asked to recognize the pairs that were studied, to increased liking. But what, exactly, does fluency
and those pairs were intermixed with new pairs of mean? Toplolinski and Strack39 tested the hypothesis
letters of which half are typed with the same finger that fluency is literally the fluency of the specific neu-
and half with different fingers. On the hypotheses that romuscular system used in producing or simulating
(1) recognition judgments reflect in part a familiarity the stimulus. In the first phase of the experiment, the
or fluency judgment, and (2) that effector-specific German participants read novel Greek words (e.g.,
motor codes are part of the memory, participants pantokrator) and listened to novel flute melodies.
should be particularly likely to mistakenly respond Producing the words requires activity in the speech
old to new pairs typed (fluently) with the different articulators (e.g., lips and tongue), whereas producing
fingers. Indeed, this was the finding, but, as predicted, the melodies requires activity in the vocal folds (as if
it was only true for expert typists, not for novice humming). While the participants were reading and
typists. Furthermore, when the experts fingers were listening, half were simultaneously moving the tongue
kept busy with a secondary task, the effect was and half were vocalizing Mm-hm (as when signal-
eliminated. Apparently, well-learned motor codes play ing assent). In the second phase of the experiment,
a role in memory even when those motor codes are participants continued moving the tongue or saying
not explicitly evoked during study. Mm-hm while they rated how much they liked old
and new words and old and new melodies. For partici-
pants moving the tongue, there was no mere exposure
UNIFYING PSYCHOLOGY: SOCIAL effect (no difference between ratings of old and new
PSYCHOLOGY AND EMOTION stimuli) for the words, but there was an effect for the
melodies. In contrast, for the participants vocalizing
The embodiment claim for social psychology is that Mm-hm there was no mere exposure effect for the
how we understand and interact with other people is melodies, but there was an effect for the words. Thus,
influenced by ones own bodily states. Williams and interfering with the development of fluency in partic-
Bargh34 provide a compelling demonstration of how ular neuromuscular systems interferes with the mere
bodily states influence social judgments and behavior. exposure effect.
They tested the hypothesis that physical warmth is Emotion is another area of research that is
metaphorically extended (as predicted by Lakoff) to associated with social psychology, but should be of
dimensions of social warmth, such as friendliness and concern to all psychologists. For example, as noted
helpfulness. This extension might occur because of above, Havas et al.28 demonstrated a link between
the association of literal warmth and social warmth emotional reactivity and language comprehension.
during early experiences with caregivers. To test the Niedenthal et al.4042 make a compelling case that
hypothesis, participants were asked to hold for a few emotions are strongly embodied. That is, emotions
moments either a cup of hot coffee or iced coffee. are not just things we think about. Instead, emotions
Then, they were asked to read a resume and rate the involve bodily changes that have a strong effect on
person on various dimensions, some of which related cognition and action.43,44
to social warmth. As predicted, holding the hot coffee As one instance of how bodily state influences
cup increased ratings on social warmth dimensions emotions in a social context, consider Oberman
but not other dimensions. In a second study, the et al.45 They tested the hypothesis that facial mimicry
participants were asked to choose a gift either for contributes to emotion recognition. That is, if A
themselves or for a friend. After holding a warm smiles, B tends to mimic that smile generating a similar
object, 54% chose the gift for a friend, whereas after state in B. This hypothesis predicts that if activity in
holding a cold object, only 25% chose the gift for a a facial muscle used in producing (and mimicking)
friend. Thus, changing the body (the literal warmth a particular emotional expression is blocked, then
of the hands) affects judgments of others as well as recognition of that emotion should be reduced. In
social behavior. Other examples of embodied effects the experiment, participants determined the emotion

Vo lu me 1, Ju ly /Au gu s t 2010 2010 Jo h n Wiley & So n s, L td. 591

 Advanced Review wires.wiley.com/cogsci

being expressed in a picture of a face. In one condition, positively correlated with empathy53 and negatively
the participants were biting on a pen using only correlated with autistic behaviors.54
the teeth in order to block smiling (cf. Ref 28). As Neuropsychological data have also played a
predicted, biting on the pen selectively reduced the role in investigating the function of MNS. For
recognition of happiness, but not fear, or sadness. example, Pazzaglia et al.55 demonstrated that patients
with lesions that produce buccofacial apraxia have
difficulty in recognizing the meaning of mouth-
UNIFYING PSYCHOLOGY: MOTOR produced sounds, but not hand-produced sounds,
RESONANCE, NEUROSCIENCE, whereas just the opposite is found for patients with
lesions that produce limb apraxia. Fazio et al.56
NEUROPSYCHOLOGY, THEORY
demonstrated that patients with aphasia due to lesions
OF MIND, AND AUTISM in Brocas area (an area previously associated with
Much of the work reviewed above is consistent with speech production) had difficulty reordering pictures
the discovery (and speculation about) mirror neurons taken from a movie of hierarchical human actions
(see Ref 46, and Ref 47 for a review). The defining (e.g., opening a door), but not pictures taken from
feature of mirror neurons is that they are active movies of equally complex physical events (e.g., a
both when an animal is engaged in a task and when bicycle falling over). That is, lesions in Brocas area,
the animal observes another engaged in the same or the human homolog of macaque area F5 where mirror
related task. With macaques, the evidence for mirror neurons were first identified, affect action recognition
neurons is very strong. That is, using an electrode to even when there is no verbal content in the task.
record from a single mirror neuron reveals that the On the basis of data such as these, Gallese et al.57
neuron is equally active when the animal is acting or suggest that mirror neurons provide a unifying base
observing another engaging in similar actions. for social cognition. That is, mirror neurons associated
Mirror neurons may indicate how motor with action provide a mechanism for mimicking
resonance can be used to recognize the actions, others (e.g, the Chameleon effect of Chartrand &
emotions, and intents of others. When A observes Bargh58 ) that leads to the understanding of the
B acting, there is resonance in As mirror neurons that others action goals. Similarly, there is evidence
are active when A takes the same action. Because A that the insula and perhaps other neural structures
knows As goal in taking that action, A can impute associated with emotional reactivity use mirror
that goal to B. If this hypothesis is correct, then mirror mechanisms. As Gallese et al.57 summarize, Social
neurons are likely to play an important role in greasing cognition is not only thinking about the contents of
the wheels of social interaction and cooperation. someone elses mind. . .Our brains, and those of other
A human mirror neuron system (MNS) may primates, appear to have developed a basic functional
play a similar role in social interaction, contributing mechanism, a mirror mechanism, which gives us an
to theory of mind and language processes, and mirror experiential insight into other minds. This mechanism
neuron functioning may contribute to the symptoms could provide the first unifying perspective of the
of autism spectrum disorder. However, research on neural basis of social cognition (page 401, emphasis
potential human MNS is more inferential, given the in the original).
relative scarcity of single cell recording (but see
Ref 48). When using functional magnetic resonance
imaging, an MNS is inferred when activity in a UNIFYING PSYCHOLOGY:
particular cortical area (as reflected in the BOLD PSYCHOPATHOLOGY AND CLINICAL
signal) is similar during action recognition and action PSYCHOLOGY
production, and when the cortical area is a likely
homolog of an area in macaque cortex in which The embodiment claim for clinical psychology is that
mirror neurons have been identified. Using this or a characteristics of the body can be used to understand
similar logic, it appears that the human MNS responds clinical symptoms. For example, Lindeman and
more to actions that the perceiver can perform than Abramson12 propose a theory of motor incapacity in
to actions the perceiver is familiar with through vision depression by combining the notions of simulation
alone (e.g., see Ref 49); the MNS is more sensitive with Lakoffs8 analysis of conceptual metaphors.
to social actions than similar nonsocial actions;50 the They begin by asking how hopelessness associated
MNS responds to both visually perceived actions and with depression leads to slow motor movements and
the linguistic description of actions;51 the MNS plays lethargy. The proposed answer has three components.
a role in speech perception;52 activity in the MNS is The first is that hopelessness is conceptualized (by the

592 2010 Jo h n Wiley & So n s, L td. Vo lu me 1, Ju ly /Au gu s t 2010

 WIREs Cognitive Science Embodiment as a unifying perspective

person feeling hopeless) as a type of motor incapacity. English) spend considerable time in learning the arbi-
That is, ones understanding of hopelessness is trary sounds of letters and how to blend those sounds.
based on (i.e., metaphorically related to in Lakoffs For most children, this is a difficult task, and even
terminology) the understanding of physical incapacity: when a child succeeds in pronouncing a word (e,g.,
Just as one cannot affect the world when physically duh.....oh....g), the pronunciation is often strained
incapacitated, one cannot affect the world when and unlike the fluid pronunciation in conversation.
hopeless. Second, when feeling hopeless, one simulates Furthermore, consider that in contrast to oral lan-
the sensorimotor experience of physical incapacity. guage learning, when the child succeeds in reading a
That is, much as understanding a sentence such as word such as dog, there are no dogs in the environ-
Anna delegated the responsibilities to you requires ment, there is no barking, there is no petting. Even
a simulation of physical giving,17 when feeling when reading in a picture book, reference to the pic-
hopeless, one simulates physical incapacity. Third, tures is haphazard at best. In other words, when a child
this simulation produces the physiological changes is learning to read, the connection between the symbol
that underlie low energy and psychomotor retardation (the written word) and the referent is often missing.
typical of hopelessness depression. What gives this For those children who do not pick up on the need
analysis force is the impressive amount of data that to forge that connection, reading becomes a boring
Lindeman and Abramson bring to bear, and their exercise in name-calling, much as an adult might find
demonstrations of how the theory can account for reading in an unknown foreign language to be boring.
extant data and make other testable predictions. On this analysis, one way of enhancing reading com-
prehension is to enforce the connection between the
symbols and their embodied meanings, and to teach
UNIFYING PSYCHOLOGY: APPLIED children how to make that connection on their own.
EDUCATIONAL PSYCHOLOGY We65,66 have developed a reading comprehen-
sion intervention, Moved by Reading, that does just
The embodiment claim for education is that the that. In the first phase, children read texts situated in
abstract symbols used in formal educationwords a particular scenario, such as a farm. In addition to
and syntax in reading, numbers and operators in
the text, in front of the child is a set of toys (e.g., a
mathneed to be grounded in bodily experience.
toy barn, corral, tractor, animals) that can be used to
A more thorough review of this approach to edu-
ground aspects of the text. In the physical manipu-
cation can be found in Ref 59 and the relation
lation condition, a child reads a sentence (e.g., The
between educational games and embodiment can be
farmer drives the tractor to the barn) and then phys-
found in Refs 6063. Lillard64 reviews the relation
ically manipulates the toys to simulate the sentence.
between embodiment and methods used in Montes-
This activity forces the child to map the words (e.g.,
sori schools. Here I will review just some of the
tractor) to objects, and to map syntactic relations
research addressing embodiment and the teaching of
reading comprehension. (e.g., who did what to whom) to their actions: it is
Why do some children hate to read when they the farmer who drives the tractor, not vice versa. In
love other uses of language such as talking, watching a control condition, children read the same texts with
movies, and being read to? There are several answers the same toys, but instead of acting out the sentences
derived from cognitive theory that almost certainly they are asked to reread them. The basic finding is
have some truth in them: some children may need that physical manipulation greatly improves compre-
extensive practice in reading to develop fluency, and hension, often with an effect size (Cohens d) of 1.0
some children may simply not have the cognitive or more. Furthermore, the effect is found for informa-
capacity (e.g., working memory capacity) to enjoy tion from the simulated sentences themselves and for
reading. Another possibility, however, is based on the information from other parts of the text.
notion of grounding symbols in bodily experience. In the second phase of Moved by Reading, the
Consider that when an infant is first exposed to toys are removed and children are taught to engage in
language, the words and their referents are frequently imagined manipulation (IM). That is, they are asked
paired. For example, a mother might say, Here is to imagine manipulating the toys. We believe that
your bottle, and literally give a bottle to the child. this IM instruction is more easily understood than
Or, when a father says, Wave bye-bye, the father other imagery instructions used with children such as
performs waving gestures. make pictures in your head. Because children have
In contrast, when children are learning to read, physically manipulated the toys, the content of what
they must (at least with an opaque orthography as in should be imagined during IM is clear to the children.

Vo lu me 1, Ju ly /Au gu s t 2010 2010 Jo h n Wiley & So n s, L td. 593

 Advanced Review wires.wiley.com/cogsci

Children in the first and second grades are able to its roots. Although students are asked to memorize
to use the IM instruction to achieve large (again, often the definition of psychology as the science of behav-
with effect sizes of 1.0 or more) gains in comprehen- ior, they are not given many ways to ground that
sion when reading new stories from the same scenario. definition. As noted by Baumeister et al.69 much of
Children in the third and fourth grades are able to use psychology has become the science of self-reports
IM both for stories from the original scenario and for and finger movements. Embodiment works toward
stories from new scenarios.67 regrounding psychology in behavior. This is not to
We believe that a similar analysis applies when say that embodiment is a renamed behaviorism in
adults are reading in novel domains (e.g., science). which it is impermissible to study processes such as
The adult must learn the meanings of new terms such attention, memory, and so on. Instead, embodiment
as force. Although a verbal definition could be mem- is a call to analyze those processes as arising from the
orized, learning the meaning in a way that is useful in recurrent, dynamic interactions of behavior, brain,
further understanding requires grounding the term in bodily processes, and changes in the physical and
bodily experiences, in this case, experiences of pushing social world. Thus, the promise of embodiment is
and being pushed (cf. Ref 68). twofold. By focusing on these interactions, behavior
will be brought back into psychology. At the same
time, this focus will provide psychologists with a
SUMMARY: PSYCHOLOGY AS THE
unifying conceptual framework.
SCIENCE OF BEHAVIOR
Work on embodiment has a long way to go to
unify psychology. Nonetheless, given that research NOTE
on embodiment has begun only recently, and given
a Why is the mere exposure effect considered to be
that it has been adopted (to date) by only a small
percentage of psychologists, progress has been swift. a phenomenon in social psychology? There is little
This rate of progress suggests that embodiment is on reason other than it was first report by Zajonc, a
the right track. social psychologist. This sort of unproductive parcel-
Another reason to suspect that embodiment is ing of phenomena is what an embodied approach to
on the right track is that it brings psychology back psychology can avoid.

ACKNOWLEDGEMENTS
This article is an expanded version of a presentation as part of the Architecture of Cognition symposium
presented at the meeting of the Association for Psychological Science in 2009. Work on this article was
supported by the National Science Foundation under Grant No. BCS 0744105 to Arthur Glenberg. Any
opinions, findings, and conclusions or recommendations expressed in this material are those of the author and
do not necessarily reflect the views of the National Science Foundation.

REFERENCES
1. Mischel W. Becoming a cumulative science. Assoc for 6. Landauer TK, Dumais ST. A solution to Platos prob-
Psychol Science Observer 2009, 22. lem: the Latent Semantic Analysis theory of acquisition,
2. Llinas R. The i of the vortex. Cambridge, MA: MIT induction and representation of knowledge. Psychol
Press; 2001. Rev 1997, 104:211240.

3. Dreyfus HL. The current relevance of Merleau-Pontys 7. Newell A. Physical symbol systems. Cogn Sci 1980,
phenomenology of embodiment. Electron J Anal Philos 4:135183.
1996, 4. 8. Lakoff G. Women, Fire, and Dangerous Things: What
4. de Vega M, Glenberg AM, Graesser AC. Symbols, Categories Reveal about the Mind. Chicago, IL: Uni-
Embodiment and Meaning. Oxford, UK: Oxford Uni- versity of Chicago Press; 1987.
versity Press; 2008. 9. Gallese V, Lakoff G. The brains concepts: the role of
5. Harnad S. The symbol grounding problem. Physica D the sesnsory-motor system in conceptual knowledge.
1990, 42:335346. Cogn Neuropsychol 2005, 22:455479.

594 2010 Jo h n Wiley & So n s, L td. Vo lu me 1, Ju ly /Au gu s t 2010

 WIREs Cognitive Science Embodiment as a unifying perspective

10. Gibbs R. Embodiment and Cognitive Science. New 29. Glenberg AM, Webster BJ, Mouilso E, Lindeman LM.
York: Cambridge University Press; 2006. Gender, emotion, and the embodiment of language
11. Boroditsky L, Ramscar M. The roles of body and mind comprehension. Emotion Rev 2009, 1:151161.
in abstract thought. Psychol Sci 2002, 13:185188. 30. Strack F, Martin LL, Stepper S. Inhibiting and facilitat-
12. Lindeman L, Abramson L. The role of simulation of ing conditions of the human smile: a nonobtrusive test
motor incapacity in depression. J Cogn Psychotherapy of the facial feedback hypothesis. J Pers Soc Psychol
2008, 22:228249. 1988, 54:768777.

13. Barsalou LW. Perceptual symbol systems. Behav Brain 31. Pecher, et al. 2005.
Sci 1999, 22:577660. 32. Brunel L, Labeye E, Lesourd M, Versace R. The sensory
14. Pecher D, Zeelenberg R, Barsalou LW. Verifying nature of episodic memory: Sensory priming effects due
different-modality properties for concepts produces to memory trace activation. J. Exp Psychol Learn Mem
switching costs. Psychol Sci 2003, 14:119124. Cogn 2009, 35:10811088.
15. Spence C, Nicholls MER, Driver J. The cost of expect- 33. Yang S, Gallo D, Beilock SL. Embodied memory judg-
ing events in the wrong sensory modality. Percept ments: a case of motor fluency. J Exp Psychol Learn
Psychophys 2000, 63:330336. Mem Cogn 2009, 35:13591365.
16. Glenberg AM, Kaschak MP. Grounding language in 34. Williams LE, Bargh JA. Experiencing physical
action. Psychonomic Bull Rev 2002, 9:558565. warmth promotes interpersonal warmth. Science 2008,
322:606607.
17. Glenberg AM, Sato M, Cattaneo L, Riggio L, Palumbo
D, et al. Processing abstract language modulates motor 35. Mussweiler T. Doing is for thinking! Stereotype acti-
system activity. Q J Exp Psychol 2008, 61:905919. vation by stereotypic movements. Psychol Sci 2006,
17:1721.
18. Zwaan RA, Taylor LJ. Seeing, acting, understanding:
motor resonance in language comprehension. J Exp 36. Jostmann NB, Lakens D, Schubert TW. Weight as
Psychol Gen 2006, 135:111. an embodiment of importance. Psychol Sci 2009,
19. Hauk O, Johnsrude I, Pulvermuller F. Somatotopic 20:11691174.
representation of action words in human motor and 37. Semin GR, Smith RE. Embodied Grounding: Social,
premotor cortex. Neuron 2004, 41:301307. Cognitive, Affective, and Neuroscientific Approaches.
20. Sommerville JA, Woodward AL, Needham A. Action New York: Cambridge University Press; 2008.
Experience Alters 3-Moth-old Infants Perception of 38. Zajonc RB. Attitudinal effects of mere exposure. J Pers
Others Actions, 2005. Soc Psychol 1968, 9:127.
21. Campos JJ, Anderson DI, Barbu-Roth MA, Hubbard 39. Topolinski S, Strack F. Motormouth: mere exposure
EM, Hertenstein MJ, et al. Travel broadens the mind. depends on stimulus-specific motor simulations. J Exp
Infancy 2000, 1:149219. Psychol Learn Mem Cogn 2009, 35:423433.
22. Smith LB. Action alters shape categories. Cogn Sci 40. Niedenthal PM. Embodying emotion. Science 2007,
2005, 29:665679. 316:10021005.
23. Glenberg AM. Language and action: creating sensible 41. Niedenthal PM, Barsalou LW, Winkielman P, Krauth-
combinations of ideas. In: Gaskell G, ed. The Oxford Gruber S, Ric F. Embodiment in attitudes, social
Handbook of Psycholinguistics. Oxford, UK: Oxford perception, and emotion. Pers Soc Psychol Rev 2005,
University Press; 2007, 361370. 9:184211.
24. Kaschak MP, Madden CJ, Therriault DJ, Yaxley RH, 42. Niedenthal PM, Winkielman P, Mondillon L, Ver-
Aveyard M, et al. Perception of motion affects language meulen N. Embodied emotion concepts. J Pers Soc
processing. Cognition 2005, 94:B79B89. Psychol 2009, 96:11201136.
25. Meteyard L, Bahrami B, Vigliocco G. Motion detec- 43. Frijda N. Emotions and action. In: Manstead ASR,
tion and motion verbs: language affects low-level visual Frijda N, Fischer A, eds. Feelings and Emotions: The
perception. Psychol Sci 2007, 18:10071013. Amsterdam Symposium. Cambridge: Cambridge Uni-
26. Meteyard L, Zokaei N, Bahrami B, Vigliocco G. Visual versity Press; 2005.
motion interferes with lexical decision on motion 44. Bradley MM, Codispoti M, Cuthbert BN, Lang PJ.
words. Curr Biol 2008, 18:R732R733. Emotion and motivation I: defensive and appeti-
27. Chambers CG, Tanenhaus MK, Magnuson JS. Actions tive reactions in picture processing. Emotion 2001,
and affordances in syntactic ambiguity resolution. 1:276298.
J Exp Psychol Learn Mem Cogn 2004, 30:687696. 45. Oberman L, Winkielman P, Ramachandran VS. Face
28. Havas DA, Glenberg AM, Rinck M. Emotion simula- to face: blocking facial mimicry can selectively impair
tion during language comprehension. Psychon Bull Rev recognition of emotional expressions. Socia Neurosci
2007, 14:436441. 2007, 2:167178.

Vo lu me 1, Ju ly /Au gu s t 2010 2010 Jo h n Wiley & So n s, L td. 595

 Advanced Review wires.wiley.com/cogsci

46. Gallese V, Fadiga L, Fogassi L, Rizzolatti G. Action 59. Glenberg AM. Embodiment for education. In: Calvo
recognition in the premotor cortex. Brain 1996, P, Gomila A, eds. Handbook of Cognitive Science:
119:593609. An Embodied Approach. Amsterdam: Elsevier Science;
47. Rizzolatti G, Craighero L. The mirror-neuron system. 2008, 355372.
Annu Rev Neurosci 2004, 27:169192. 60. Barab S, Zuiker S, Warren S, Hickey D, Ingram-Goble
48. Iacoboni M. Imitation, empathy, and mirror neurons. A, et al. Situationally embodied curriculum: relating
Annu Rev Psychol 2009, 60:653670. Formalisms and contexts. Sci Educ 2007, 91:133.
49. Calvo-Merino B, Grezes J, Glaser DE, Passingham RE, 61. Gee JP. What Video Games have to Teach us About
Haggard P. Seeing or doing? Influence of visual and Learning. New York: Palgrave; 2003.
motor familiarity in action observation. Curr Biol 2006, 62. Gee JP. Good Video Games + Good Learning: Col-
16:19051910. lected Essays on Video Games, Learning and Literacy
50. Kilner JM, Marchant JL, Frith CD. Modulation of the New York: Peter Lang; 2007.
mirror system by social relevance. Social Cogn Affective 63. Gee JP. Learning and Games. In: Salen K, ed. The
Neurosci 2006, 1:143148. Ecology of Games: Connecting Youth, Games, and
51. Tettamanti M, Buccino G, Saccuman MC, Gallese V, Learning, The John D. and Catherine T. MacArthur
Danna M, et al. Listening to action-related sentences Foundation Series on Digital Media and Learning.
activates fronto-parietal motor circuits. J Cogn Neu- Cambridge, MA: The MIT Press; 2008.
rosci 2005, 17:273281. 64. Lillard AS. Montessori: The Science Behind the Genius.
52. DAusilio A, Pulvermuller F, Salmas P, Buflari I, New York: Oxford University Press; 2005.
Begliomini C, et al. The motor somatotopy of speech 65. Glenberg AM, Gutierrez T, Levin JR, Japuntich S,
perception. Curr Biol 2009, 19:381385. Kaschak MP. Activity and imagined activity can
53. Gazzola V, Aziz-Zadeh L, Keysers C. Empathy and the enhance young childrens reading comprehension.
somatotopic auditory mirror system in humans. Curr J Educ Psychol 2004, 96:424436.
Biol 2006, 16:18241829. 66. Glenberg AM, Goldberg A, Zhu X. Improving early
54. Dapretto M, Davies MS, Pfeifer JH, Scott AA, Sigman reading comprehension using embodied CAI. Instruc-
M, et al. Understanding emotions in others: mirror tional Sci. In press.
neuron dysfunction in children with autism spectrum
67. Glenberg AM, Jaworski B, Rischal M, Levin JR. What
disorders. Nat Neurosci 2006, 9:2830.
brains are for: action, meaning, and reading compre-
55. Pazzaglia M, Pizzamiglio L, Pes E, Aglioti SM. hension. In: McNamara D, ed. Reading Comprehension
The sound of actions in apraxia. Curr Biol 2008, Strategies: Theories, Interventions, and Technologies.
18:17661772. Mahwah, NJ: Lawrence Erlbaum Associates; 2007,
56. Fazio, et al. 2009. 221240.
57. Gallese V, Keysers C, Rizzolatti G. A unifying view of 68. Talmy L. Force dynamics in language and cognition.
the basis of social cognition. Trends Cogn Sci 2004, Cogn Sci 1988, 12:49100.
8:396403. 69. Baumesiter RF, Vohs KD, Funder DC. Psychology as
58. Chartrand TL, Bargh JA. The chameleon effect: the the science of self-reports and finger movements: what-
perception-behavior link and social interaction. J Pers ever happened to actual behavior? Perspect Psychol Sci
Soc Psychol 1999, 76:893910. 2007, 2:396403.

FURTHER READING
Barsalou LW. Grounded cognition. Annu Rev Psychol 2008, 59:617645.
Calvo P, Gomila A, eds. Handbook of Cognitive Science: An Embodied Approach. Amsterdam: Elsevier Science;
355372.
Semin GR, Smith RE, eds. Embodied Grounding: Social, Cognitive, Affective, and Neuroscientific Approaches.
New York: Cambridge University Press; 4370.
Shapiro L. The embodied cognition research program. Philos Compass, 2007, 2:338346.
Umilta MA, Escola L, Intskirvell I, Grammont F, Rochat M, et al. When pliers become fingers in the monkey
motor system. Proc Natl Acad Sci U S A 2008, 105:22092213.
Umilta MA, Kohler E, Gallese V, Fogassi L, Fadiga L, et al. I know what you are doing: a neurophysiological
study. Neuron 2001, 31:155165.
de Vega M, Glenberg AM, Graesser AC. Symbols, Embodiment and Meaning. Oxford, UK: Oxford University
Press; 2008.
Wilson M. Six views of embodied cognition. Psychon Bull Rev 2002, 9:625636.

596 2010 Jo h n Wiley & So n s, L td. Vo lu me 1, Ju ly /Au gu s t 2010