3.carnie Chapter1 Introduction

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/287227101
On the need for embodied and disembodied cognition
Article in Frontiers in Psychology · January 2011
CITATIONS READS
23 172
1 author:
Guy O. Dove
University of Louisville
21 PUBLICATIONS 605 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Book on Abstract Concepts and Grounded Cognition View project
All content following this page was uploaded by Guy O. Dove on 17 September 2017.
The user has requested enhancement of the downloaded file.

HYPOTHESIS AND THEORY ARTICLE
published: 25 January 2011
doi: 10.3389/fpsyg.2010.00242
On the need for embodied and dis-embodied cognition

Guy Dove*
Department of Philosophy, University of Louisville, Louisville, KY, USA
Edited by: This essay proposes and defends a pluralistic theory of conceptual embodiment. Our concepts
Anna M. Borghi, University of Bologna, are represented in at least two ways: (i) through sensorimotor simulations of our interactions
Italy
with objects and events and (ii) through sensorimotor simulations of natural language processing.
Reviewed by:
Linguistic representations are “dis-embodied” in the sense that they are dynamic and multimodal
Bradford Mahon, University of
Rochester, USA but, in contrast to other forms of embodied cognition, do not inherit semantic content from this
Jesse Prinz, City University of New embodiment. The capacity to store information in the associations and inferential relationships
York, USA among linguistic representations extends our cognitive reach and provides an explanation of our
*Correspondence: ability to abstract and generalize.This theory is supported by a number of empirical considerations,
Guy Dove, Department of Philosophy,
including the large body of evidence from cognitive neuroscience and neuropsychology
University of Louisville, 313B
Humanities Building, Louisville, KY supporting a multiple semantic code explanation of imageability effects.
40292, USA. Keywords: concepts, imageability, embodied cognition, language, representation, symbol grounding
e-mail: [email protected]
INTRODUCTION clearly demonstrated by the heated nature of the debate concern-

In this essay, I propose and defend a new take on a familiar idea. The ing the possibility that analog perceptual representations might be
familiar idea is that our concepts are encoded in at least two gen- employed in mental imagery tasks (Pylyshyn, 1973, 1981; Kosslyn
eral types of semantic representations: one type that is perception and Shwartz, 1977). Now, there is general agreement that behavioral
and motor based and another that is language based (Paivio, 1971, and neural evidence suggests that mental imagery (Kosslyn, 1994)
1986). Although most concepts employ both types of representa- and motor imagery (Jeannerod, 1995; Grèzes and Decety, 2001)
tions, abstract concepts tend to depend more on linguistic repre- depend on sensory and motor representations respectively.
sentations than concrete concepts do. What separates my version Within the last two decades, a growing number of researchers
of this idea from most previous ones is that I develop it within an and philosophers have argued that cognitive science needs to reo-
embodied approach to cognition (although see Barsalou et al., 2008; rient itself with respect to its fundamental assumptions about the
Louwerse and Jeuniaux, 2008 for related yet distinct views). nature of mind and cognition. These researchers and philosophers
My defense of this new take has three parts. The first part out- contend that cognitive processes need to be viewed as fundamen-
lines and motivates an embodied approach to concepts based on tally based in our bodily interactions with the world. Clark (1998,
simulation. The second part examines a challenge that faces any p. 506) expresses this view clearly in his economical assertion that,
form of embodied cognition: the problem of abstraction. After “Biological brains are first and foremost the control systems for
making the observation that the symbolic structure of language is biological bodies.” The idea is that we cannot hope to understand
well suited to solving this problem, I propose that language should the functioning of the brain without appreciating the central role
be seen as a form of what I refer to as “dis-embodied” cognition. it plays in guiding perception and action. This view has lead to a
What I mean by this is that linguistic representations are embod- robust and diverse research program in which investigators examine
ied in the neurophysiological sense that they rely on sensorimotor the possible ways in which thinking, remembering, and under-
simulation but, unlike other embodied forms of cognition, they standing language are shaped by the fact that we dynamically inter-
do not inherit semantic content from this fact. They do, however, act with our complex physical and social environment by means
accrue semantic content through their associations and inferential of perceptual and motor capacities (Wilson, 2002). Embodied
relationships with other linguistic representations. The third part theories of cognition often suggest that concepts are understood
surveys empirical evidence that supports the existence of separate via sensorimotor simulations. Neural systems that are involved in
semantic codes. understanding real objects, actions, and events in the world are
used to internally simulate those objects, actions, and events at
EMBODIED CONCEPTS later points in time.
Historically, cognitive scientists have presumed that higher cogni-
tive processes are carried out by computations involving amodal THE THEORETICAL PROMISE OF EMBODIED CONCEPTS
mental representations (i.e., representations that are not located Within cognitive science, the orthodox approach to concepts views
within a sensorimotor modality). The precise nature of these rep- them as containing amodal representations. This approach posits
resentations was a matter of some debate. For instance, a great deal mental symbols that are manipulated solely based on their syntactic
of controversy has surrounded the issue of how language-like they properties. By assumption, there is no intrinsic connection between
might be (Fodor, 1975). The presumption of amodality, however, these symbols and what they represent. This approach faces a well-
went largely unquestioned. The strength of this presumption was known challenge: the symbol grounding problem. Harnad (1990,
www.frontiersin.org January 2011 | Volume 1 | Article 242 | 1

Dove Embodied and dis-embodied cognition
p. 335) summarizes this problem with the question, “How can the (2005) identifies two distinct strains of embodied cognition – one
meanings of the meaningless symbol tokens, manipulated solely on that focuses on affordances and situated action and the other that
the basis of their (arbitrary) shapes, be grounded in anything but focuses on simulation – and argues that both are true depend-
other meaningless symbols?” Perhaps the easiest way to think of ing on the context. I am going to limit myself to the simulation
this problem is to imagine trying to learn a foreign language from framework here for a couple of reasons. The first is that I believe
a dictionary in that language. Each word would be defined in terms that this framework is more flexible than critics assume. An unfor-
of its connections to other words. In order to avoid this problem, tunate consequence of Barsalou’s use of the term perceptual symbol
the meaning of at least some mental symbols must be grounded is that it gives the false impression that simulations are based in
in something other than their syntactic properties. perception and not in action mechanisms. However, nothing in
A key impetus for the hypothesis that concepts are couched in the theory prevents purely motor-based simulations. Perceptual
sensorimotor representations is the belief that this will help with the symbols are thus compatible with, for example, action schemas
symbol grounding problem. In order to see just how it might help, (Glenberg and Robertson, 1999). The second reason is that I am
we need to have a clear conception of what an embodied account committed to a representational approach to concepts. One of the
of concepts might look like. For that reason, I am going to briefly issues that separate different views of embodiment is the status
sketch what I take to be the strongest and most developed embodied of representations. Many proponents of affordances and situated
accounts of concepts: the perceptual symbol theory (Barsalou, 1999). action embrace non-representational accounts of cognition – often
I should emphasize, though, that many of the points made in this appealing to the promise of dynamical systems theory. Based largely
essay extend beyond this particular theory and do not depend on its on this issue, Clark (1997) distinguishes between embodied cogni-
ultimate success. A core tenet of perceptual symbol system theory is tive science and radical embodied cognitive science. Clark (1997,
that sensorimotor simulations of experience are of central impor- 2008) defends the former while theorists such as Chemero (2009)
tance to our concepts. Intuitively, the idea is that our conceptualiza- defend the latter. Siding with Clark, I assume that the notion of
tion of a category consists of simulating the experience of perceiving representation is too useful to give up and, furthermore, that an
and/or acting on exemplars of that category. Such simulations are empirically successful theory of concepts will involve an appeal to
the result of a kind of neurophysiological re-enactment: information representations (Markman and Dietrich, 2000).
concerning the neural activation patterns associated with perception A perceptual symbol consists of a neurophysiological re-
or action, which has been captured and stored by conjunctive neu- enactment of a collection of sensorimotor representations. It
rons in neighboring association areas or convergence zones (Damasio can be thought of as having perceptual content because there are
and Damasio, 1994), is used later in absence of relevant input to certain states of affairs in the world that would be likely to elicit
generate a partial reactivation of the sensorimotor representations. these representations under normal conditions. Barsalou argues
Perceptual symbols have a number of properties that make them that this perceptual content can provide a leg up with regard to
well suited to serve as conceptual representations (Barsalou, 1999, intentional content. He writes (Barsalou, 1999, p. 597; emphasis
2003). First, simulations need not be conscious – that is, they may in the original):
contain unconscious perceptual representations (for evidence to
Where perceptual symbols do have an advantage [over amodal
this effect see Pecher et al., 2009). This property removes some of the symbols] is in the ability of their content to play a heuristic role in
traditional objections to imagistic theories of cognition that turn establishing reference. Although perceptual content is rarely defini-
on the unreliability or vagueness of introspection. Second, simula- tive for intentionality, it may provide a major source of constraint
tions will often be schematic in the sense they contain only some of and assistance in determining what a symbol is about.
the sensorimotor representations involved in the experience being
simulated. For instance, a simulation in the visual modality of the The general idea is that perceptual symbols help us refer to objects
concept DOG might involve shape representations but not color and events because they are already causally connected with those
representations. Third, they will typically be multi-modal in the objects and events. This causal connection does not fully determine
sense that they involve the reactivation of perceptual representations the conceptual content of a perceptual symbol but it can help secure
in several modalities. Fourth, perceptual symbols provide a novel that content.
means of drawing the type/token distinction (Barsalou, 1999, 2003). Although embodied cognition has promise with respect to
This is achieved through distinguishing simulators and simulations. helping with the symbol grounding problem, it seems too early to
A simulator is a distributed system spanning association and sen- declare victory for two reasons. The first is that it is not clear that
sorimotor areas. To possess a concept, such as DOG, is to have a the problem has been fully solved (Taddeo and Floridi, 2005). The
skill or ability to generate appropriate perceptual representations second is that other approaches may have the conceptual resources
of dogs in a given situation. An innovative aspect of Barsalou’s to address the problem. Instead of proclaiming that embodiment
account is that it holds that these simulations are context-sensitive: solves this longstanding problem, I am going to make a weaker
simulations for a given concept vary depending on the context and and hopefully less controversial claim: the heuristic role identified
the speaker’s goals. For example, they might represent objects from by Barsalou is an attractive design feature of perceptual symbols.
a particular perspective. Typically, simulations will involve only a A conceptual system containing perceptual symbols can benefit
small subset of the information stored in memory. from the role that sensorimotor representations play in guiding
Although I believe that simulation-based accounts of embodi- action and perception. To be more precise, I am going to claim
ment have the most empirical promise, I should acknowledge that that this design feature is more beneficial with some concepts than
there are other theoretical conceptions of embodiment. Borghi it is with others.
Frontiers in Psychology | Cognition January 2011 | Volume 1 | Article 242 | 2

EMPIRICAL EVIDENCE Using a different experimental paradigm, Buccino et al. (2005)

There is little question that embodied cognition has been a produc- found that listening to action-related sentences modulated activity
tive research program. New research seems to emerge daily. Due in the motor system. Motor evoked potentials (MEPs) recorded
to this abundance, I am only going to offer a selective review. My from hand and foot muscles were specifically modulated by hand-
purpose is not to be comprehensive but, instead, to provide general related and foot-related action sentences respectively.
motivation for an embodied approach to concepts. More support is provided by the fact that lesions can lead to
A number of behavioral experiments support the notion that the loss of multiple categories that share perceptual properties
perceptual representations are central to some cognitive tasks. (Simmons and Barsalou, 2003). For instance, Adolphs et al. (2000)
For instance, Pecher et al. (2003) found a modality-switching cost found that damage to the somatosensory cortex was correlated with
in a linguistic task. Participants verified verbally expressed facts deficits in the visual recognition of facial expressions. The authors
involving one modality, such as the fact that leaves rustle, more propose that simulation of producing facial expressions is involved
rapidly after verifying a fact involving the same modality, such as in the recognition of facial expressions in others. A selective deficit
the fact that blenders make noise, than after verifying a fact involv- in action word processing has been found patients with motor neu-
ing a different modality, such as the fact that cranberries are tart. ron disease (Bak et al., 2001). A word of caution is needed, though,
More recently, van Dantzig et al. (2008) found a similar modality- because modality-specific damage does not explain the category-
switching cost between a perceptual detection task and a property specific deficits of all patients (Caramazza and Mahon, 2006).
verification task. Stanfield and Zwaan (2001) asked participants to A body of brain imaging data support an embodied approach
affirm whether or not pictures depicted the actions described in to concepts. Martin et al. (1996), for example, found increased
previously presented sentences. The actions had either a vertical or activation in visual areas with categories that appear to rely heavily
horizontal orientation (such as driving a nail into a floor or into a on visual information for identification. Simmons et al. (2007) find
wall). Participants responded more quickly to the pictures that had evidence of a common neural substrate for color perception and
the same orientation as the action described. Stanfield and Zwaan verification of object-associated color (e.g., taxi-yellow). Using a
(2001) suggest that the subjects generated a perceptual image of visual naming task, Chao and Martin (2000) found increased activ-
the action described in the sentence and then used this image to ity in motor areas with highly manipulable objects when compared
carry out the affirmation task. to less manipulable objects. Hauk et al. (2004) had participants
Other behavioral studies demonstrate the degree to which cog- read individual words that referred to actions involving leg, arm,
nitive tasks can be interwoven with action. For instance, Borghi and head movements such as lick, pick, and kick. They found that
et al. (2004) found a compatibility effect associated with language reading each type of action word produced increased activation
processing and action. Participants were instructed to decide in successively in the regions of M1 associated with performing
whether or not a word that followed a sentence named a part of the relevant movements. In a task where participants listened to
the object mentioned in the sentence. Half of the selected parts were action-related sentences, Tettamanti et al. (2005) found increased
found in the upper-portion of the object and half were found in activation in effector-specific premotor and motor areas.
the lower-portion. The experimenters found that responses were In sum, a number of studies using different experimental para-
faster when the direction of the key press movement (upward or digms and techniques implicate sensorimotor representations in
downward) matched the part location (upper or lower). Further various cognitive tasks. Positing perceptual symbols provides an
studies indicate that the motor representations elicited by the cogni- economical and robust explanation for a diverse set of observed
tive tasks can exhibit somatotopic specificity. For instance, Scorolli phenomena, including reaction times, the functional character of
and Borghi (2007) asked their participants to judge whether or some neuropathologies, and neural activation patterns in response
not simple sentences containing a verb and a noun were sensible to certain cognitive tasks.
or not and respond either by pressing a pedal or speaking into a
microphone. The verbs in the sentences referred to actions that were CHALLENGES TO THE EVIDENCE
typically performed with the mouth, hands, or the feet. Response Aside from the problem of abstraction, which will be discussed
times with the microphone were fastest with “mouth-sentences” in the next section, the inference to embodied cognition from the
and response times with the pedal were fastest with “foot-sentences” available evidence faces two major challenges. The first concerns
(see also Scorolli et al., 2009). how the debate is framed. Machery (2007, 2010) argues that amo-
Researchers have produced evidence using transcranial mag- dal theories are not monolithic, and there are conceivable amodal
netic stimulation (TMS) that provides compelling support for the systems that would fit with the available evidence. In a similar vein,
behavioral findings. Pulvermüller et al. (2005) carried out a TMS Mahon and Caramazza (2008) contend that the activity in senso-
study in which they found that stimulation over motor areas affects riomotor areas observed in many experiments could be the result
action word processing. They weakly stimulated different parts of of spreading activation from amodal representations. The ability
the motor system while participants performed a lexical decision to offer amodal explanations for the available evidence undermines
task on arm- and leg-related action words. Weak stimulation of left some of the hyperbolic rhetoric used by supporters of embodied
hemisphere areas associated with arm-movement led to an increased cognition. Too often, such supporters claim that the empirical
response time with arm-related words in comparison with leg-re- predictions of embodied and amodal approaches sharply diverge.
lated words, and the reverse pattern occurred with weak stimulation What Machery and Mahon and Caramazza demonstrate is that the
of motor areas associated with leg-movement. Response times were empirical decision between the embodied and amodal approaches
not modulated in a control condition with a sham stimulation. may be more difficult than some have advertised. This point seems

well taken; the issue will ultimately be decided by which approach is Dove, 2009). This is problematic because, although it is not difficult
best supported by the evidence. The defeasible position of this paper to imagine how embodiment might help us acquire concrete con-
is that the available evidence favors an embodied approach. cepts, it is difficult to see how it can be anything but a hindrance with
The second challenge is that the neuroimaging evidence does not abstract concepts such as DEMOCRACY, ELECTRON, ENTROPY,
exclude the presence of amodal representations. Indeed, many of the JUSTICE, NUMBER, PATIENCE, and TRUTH. Representations
cited imaging experiments find modulation of activity in multiple grounded in sensorimotor systems do not seem to be well suited to
brain areas. Several commentators (e.g., Weiskopf, 2007; Chatterjee, representing abstract intentional contents. For this reason, abstract
2010; Machery, 2010) point out that a number of the neuroimag- concepts remain a critical issue for embodied cognition. More is at
ing studies cited in support of embodied cognition actually find stake than simply the reach of this approach. For instance, Mahon
modulated activity in brain areas that are near – but not identical and Caramazza (2008, p. 60) use the challenge posed by abstract
to – areas used for perceptual and motor processing. This is a seri- concepts to support a parsimony argument in support of an amodal
ous challenge to a philosophical position known as neo-empiricism approach to concepts:
(Prinz, 2002). A core tenet of this position is that all conceptual
Given that an embodied theory of cognition would have to admit
representations are modality-specific (Machery, 2010). Against this ‘disembodied’ cognitive processes in order to account for the rep-
universal claim, evidence suggesting that some conceptual represen- resentation of abstract concepts, why have a special theory just for
tations are located within areas outside of areas used for perceptual concepts of concrete objects and actions?
and motor processing is damning. It is not clear, though, that such
evidence undermines a simulation-based embodied approach. While I am not convinced that such parsimony arguments have
On some level, the distributed activation patterns found in the much force (the history of psychology is rich with highly eco-
literature fit with the theory of perceptual symbols. Barsalou (2003) nomical failed theories), the core premise of this argument – i.e.,
proposes that long-term memory integration processes underlie the that abstract concepts require disembodied cognition – needs to
ability create appropriate simulations. Such processes are needed to be examined.
explain our ability to generalize and abstract away from particular
exemplars and generate the right simulations on a given occasion. THREE EMBODIED APPROACHES TO ABSTRACT CONCEPTS
This move offloads significant aspects of conceptualization into Supporters of embodied concepts have begun to address the prob-
non-perceptual association areas or convergence zones (Damasio lem of abstraction. Three main approaches exist in the literature
and Damasio, 1994). It also raises the question of whether or not (for a review see Glenberg et al., 2008). Although each approach
these areas contain amodal symbols. Barsalou et al. (2003, p. 87) has some empirical support, there are reasons to believe that
concede that “{conjunctive neurons in convergence zones consti- these approaches do not provide a full solution to the problem
tute a somewhat amodal mechanism for capturing and re-enacting of abstraction.
modality-specific states” but then go on to point out that alterna- The first and most well established approach involves meta-
tive explanations of the activity of these neurons are available that phoric extension. This approach originally emerged from work
do not require amodal symbols. They then suggest that we should in cognitive linguistics (Lakoff and Johnson, 1980; Lakoff, 1987).
pragmatically assume that convergence zones do not contain amo- The core idea is that we often understand one conceptual domain
dal symbols until evidence suggests otherwise. metaphorically in terms of another. Often, these metaphors are
This is not a satisfying solution to the challenge posed by acti- shaped by image schemas formed from our bodily interactions,
vation in convergence zones because it is provisional and ad hoc. linguistic experience, and historical context. For instance, the con-
Fortunately, there is a better way to meet this challenge: we can cept of ARGUMENT may be understood in terms of the concept
adopt a more liberal definition of an embodied concept. The fun- of WAR. The primary evidence for this approach is our use of lin-
damental intuition behind the embodied approach is that cognition guistic metaphors. Some recent behavioral studies, though, provide
is fundamentally integrated with perceptual and motor systems. evidence of the metaphorical use of space to represent abstract
Such integration does not in and of itself exclude supramodal or concepts. For instance, Boroditsky and colleagues (Boroditsky and
even amodal representations as long as the function of these rep- Ramscar, 2002; Casasanto and Boroditsky, 2008) provide evidence
resentations is to engage appropriate simulations and not to act as that some temporal judgments rely on spatial representations.
independent conceptual representations. I would even go further Richardson et al. (2003) attempted to ascertain whether or not
and suggest that the very modal/amodal distinction fits poorly comprehending abstract verbs, such as argue and respect, auto-
with an integrated embodied perspective because it presupposes matically activates spatial image schemas with a specific orientation
a clean distinction between cognition and perception. From an (horizontal for argue and vertical for respect). Participants listened
embodied perspective, no such clean distinction exists. If I am right, to short sentences while engaged in either a visual discrimination
then evidence of relevant neural activity in areas near to, but not task or a picture memory task. Reaction times suggest that there
directly associated with, a particular sensorimotor modality is not was an interaction between the horizontal/vertical orientation of
unequivocally incompatible with an embodied approach. the image schema and the horizontal/vertical orientation of the
visual stimuli.
THE PROBLEM OF ABSTRACTION The second approach is similar in spirit to the first but focuses
A well-known limitation of the evidence for embodied concepts on the importance of action schemas (Glenberg and Robertson,
is that it primarily involves concrete or highly imageable concepts 1999). The core idea of this approach is that some abstract language
(Pezzulo and Castelfranchi, 2007; Louwerse and Jeuniaux, 2008; is grounded in motor processes. A primary source of evidence is

the action–sentence compatibility effect or ACE (Glenberg and RESPECT. This ubiquity raises the question of whether a non-
Kaschak, 2002). Glenberg and Kaschak found that reaction times metaphoric understanding of the target concept is needed to anchor
decreased when response direction (a button press either away/ these metaphoric uses.
toward the body) and the implied direction of either concrete action Although the action schema approach is similar in spirit to
sentences (e.g., Andy gave you the pizza/You gave Andy the pizza) the metaphorical projection approach, it enjoys some advantages
or abstract transfer sentences (e.g., Liz told you a story/You told Liz over the metaphorical projection approach. For one, the evidence
a story) matched. They suggest that the ACE is the result of com- offered in support of this approach seems more substantial and less
petition for resources by the motor planning associated with the equivocal. Second, the developmental picture behind this approach
action and the language processing associated with the sentence. seems more plausible. It fits with the developmental evidence sug-
Adding to the behavioral research, Glenberg et al. (2008) recently gesting that concrete or highly imageable event words are easier
provide neurophysiological evidence that comprehension of both for young children to acquire than abstract ones (Maguire and
object-transfer and abstract-transfer sentences modulates motor Dove, 2008). Despite these advantages, the action schema approach
system activity. faces some of the same challenges as the metaphoric projection
The third approach proposes that, contrary to our intuitions, approach. For instance, the apparent representational flexibility of
some abstract concepts involve situated simulations (Barsalou, action schemas raises the question of how it is possible to acquire
1999). This approach is supported by evidence from feature gen- the relevant abstract concepts. If the same action schema underlies
eration experiments. In a preliminary study, Barsalou and Wiemer- various concepts, how are the differences between these concepts
Hastings (2005) asked participants to generate typical properties for represented? Another problem is that it is difficult to imagine how
three abstract concepts (TRUTH, FREEDOM, and INVENTION), action schemas can account for all abstract concepts. For instance,
three concrete concepts (BIRD, CAR, and SOFA) and three inter- it is not clear how they might handle concepts such as ELECTRON,
mediate concepts (COOKING, FARMING, and CARPETING). NUMBER, and TRUTH.
The authors report two core findings: that participants generated Finally, consider the situated simulation approach. The
situational properties with both concrete and abstract concepts and body of evidence cited in support of it is admittedly quite thin.
that participants tended to generate more event and introspective More importantly, this evidence may not resolve the issue of
properties with abstract concepts. They propose that abstract and the embodiment of conceptual representations. A supporter of
concrete concepts are generally associated with different aspects of amodal symbols could well argue that disembodied symbols
situations: abstract concepts tend to focus on social aspects while are needed to account for our ability to represent the social
concrete concepts tend to focus on physical entities and actions. In and relational aspects of situations. In the end, the most seri-
a more fully realized experiment employing similar methodology, ous problem facing the situated simulations proposal is that a
Wiemer-Hastings and Xu (2005) found that participants tended particular abstract concept such as DEMOCRACY is not likely
to produce fewer entity properties, more introspective properties, to be associated with a simple set of sensorimotor experiences
and more relational properties with abstract concepts than with (Dove, 2009).
concrete concepts. In sum, current attempts to offer an embodied solution to the
How promising are these approaches? Let us consider each problem of abstraction appear suffer from two weaknesses: insuf-
in turn. There are a number of reasons to be skeptical of meta- ficiency and incompleteness. The approaches appear to be insuffi-
phorical projection solution to the problem of abstraction. First, cient because they do not provide a full explanation of the concepts
there are reasons to question the force of the linguistic evidence to which they apply. They appear incomplete because they do not
supporting this approach. It is just not clear that such linguistic seem to capture all abstract concepts. This is not to say that these
patterns directly reflect conceptual structure. Indeed, alternative proposals have no merit. Instead, I suggest that each has some
explanations of metaphors that do not require positing metaphoric promise and empirical support, but, ultimately, more is needed to
representations are available (Murphy, 1997). Another problem is explain our ability to abstract and generalize.
that this proposal seems developmentally implausible (Murphy,
1996). For example, it seems unlikely that an understanding of the DIS-EMBODIMENT
complexities of war is required for the acquisition of the concept Supporters of an embodied approach to concepts tend to treat the
of an argument. Furthermore, evidence suggests that children’s problem of abstraction as a collection of exceptions. The task then
understanding of metaphor remains quite poor before the ages of becomes to explain a subset of these exceptions using the theoretical
8–10 (Winner et al., 1976). Finally, there is an inherent difficulty techniques and experimental designs of the research program of
faced by the attempt to capture conceptual content in terms of embodied cognition. This effort ignores the fact that abstraction
metaphor: while a metaphor enables us to highlight the similarities represents a general problem for embodied concepts. What we need
between two concepts, it cannot capture the important differences. to explain is our ability to go beyond embodied experience. Earlier
Arguments, after all, are not really wars. Recognizing the appro- we emphasized how grounding our concepts in action and percep-
priate connections between a perceptual experience and what it is tion systems may help us acquire conceptual content. Now, we need
being metaphorically extended to cover seems to require a prior to acknowledge that such grounding has potential costs associated
understanding of the concept. Without such an understanding, it with it. In particular, sensorimotor simulations seem ill-suited for
is difficult to see how one can arrive at a correct interpretation of representing conceptual content that is not closely tied to particular
a metaphor. The very ubiquity of spatial metaphor undermines experiences. The problem is that some concepts appear to require
its potential for representing a specific abstract concept such as what we might call ungrounded representations.

The orthodox position within cognitive science, clearly expressed held up as a paradigmatic example of an amodal symbol system.
in the quote from Mahon and Caramazza given above, is that such Three design features are particularly important. The first is the
“disembodied” concepts require amodal representations. If we look inherent representational arbitrariness of words and morphemes.
at the general features of the proposed embodied solutions to the There is, for example, no intrinsic similarity or other extralinguistic
problem of abstraction – particularly the metaphor projection connection of the English word cat to the category of cats. Indeed,
and action schema approaches – a different theoretical possibil- other languages associate phonetically and graphemically different
ity emerges. Each of these approaches proposes ways in which words with the same category. Furthermore, the phonemic similar-
embodied representations associated with a certain experiential/ ity of cat to cap carries no weight with respect to the contents of
cognitive domain can be used to refer to objects and events out- these words. The second is its stimulus-independence (Chomsky,
side of that domain. To capture this idea, I am going coin a new 1966). Competent speakers are able to produce linguistic utterances
term: dis-embodiment. A mental symbol is dis-embodied if (1) it in a self-generated fashion that is not an immediate response to
is embodied but (2) this embodiment is arbitrarily related to its proximal environmental stimulation. The third is its systematicity
semantic content. In other words, a mental symbol is dis-embodied (Fodor, 1975; Fodor and Pylyshyn, 1988; Pinker, 1994). The ability
if it involves sensorimotor simulations of experiences that are not to produce a sentence such as Joni loves Chachi seems to come hand
associated with its semantic content. The dash in the middle of this in hand with the ability to form other sentences such as Chachi
term is intended to distinguish this notion from the more general loves Joni and Jenny loves Chachi, etc. A common explanation of
notion of disembodiment to which Mahon and Caramazza appeal. these design features is that natural language amounts to a syntacti-
What I want to suggest is that the proposals outlined above are on cally recombinable symbol system. While there are disagreements
the right track, but they fail to provide a general solution to the concerning the cognitive architecture that underlies our linguistic
problem of abstraction. Below, I argue that natural language itself competence, a large body of linguistic research suggests that the
serves as a form of dis-embodied cognition and plays an extensive morphosyntactic structure of language is at least characterizable
role in enabling us to acquire and use abstract concepts. in terms of a productive grammar.
Now the mere fact that natural language is stimulus-independ-
LANGUAGE AS A FORM OF DIS-EMBODIED COGNITION ent and systematic does not sufficiently distinguish it from garden
One way to approach the problem of abstraction is to scrutinize variety perceptual symbols. One of the achievements of perceptual
the abstract/concrete distinction (Scorolli, 2009). A number of symbol theory is that it demonstrates how a simulation-based sym-
researchers suggest that there are qualitative differences between bol system might have these properties (Barsalou and Prinz, 1997).
abstract and concrete concepts. For example, Barr and Caplan Stimulus-independence and systematicity alone cannot establish an
(1987) propose that a meaningful distinction can be drawn between advantage of verbal over non-verbal representations with respect to
categories that are primarily represented by “extrinsic” features abstract contents. Natural language must bring something else to
(those associated with relations between two or more entities) and the table. In a philosophical exploration of possible conceptions of
those that are represented by “intrinsic” features (those associated animal and human cognition, Camp (2009) suggests that we should
with individual entities). Based on property generation studies, view stimulus-independence and recombinability as degree proper-
Wiemer-Hastings and Xu (2005) propose a two-factor account ties. She then argues that natural language enhances these features
in which abstract concepts are both less contextually specific and in at least four ways. First, natural language is likely to increase the
predominately associated with social aspects of situations. Crutch range of thoughts that any one individual may entertain because it
and Warrington (2005) propose a qualitative distinction in which enables one to hear the thoughts of others. Second, natural language
concrete concepts are organized primarily around similarity and makes it easier to reproduce the same thought in different situa-
abstract concepts are organized around semantic association. A tions because of its lack of context-sensitivity. Third, the manifest
recent eye-tracking experiment suggests that these representational syntactic structure of natural language highlights the potential
differences emerge during on-line word-recognition (Deñabeitia recombinality of thoughts and thus encourages us to entertain a
et al., 2009). Participants were presented with visual displays that wider of thoughts. Finally, natural language provides a sufficiently
included a target picture of item that was a semantic associate of rich expressive medium to allow one to represent truth-values and
an abstract or concrete word. Their eye-movements were recorded inferential relations among thoughts. These enhancements mean
as they listened to the relevant words. They tended to fixate more that a creature with language is likely to enjoy a general cognitive
(and earlier) on depicted objects that were associates of abstract advantage over a creature that does not.
words than associates of concrete words. Overall, evidence of a A primary benefit afforded by a natural language is that it
qualitative distinction between abstract and concrete concepts is provides a representational system that can play the integrative
growing. What is the source of this distinction? I propose that it role traditionally associated with amodal symbols. Consider the
arises from an asymmetry between the types of representations following argument for the necessity of amodal representations.
employed by abstract and concrete concepts. While concrete After recognizing the existence of independent sensorimotor codes,
concepts generally depend on both linguistic and non-linguistic Jackendoff (1992, p. 3) contends that amodal representations are
perceptual symbols, abstract concepts tend to rely primarily on necessary because “{none of these forms of input and output
linguistic perceptual symbols. information suffices to explain the way that we understand the
Natural language has a number of design features commonly world in terms of objects, their motions, our actions on them,
associated with amodal symbol systems that make it well suited to and so forth.” The general idea is that amodal representations are
representing abstract concepts. Indeed, natural language is often needed to capture generalizations about entities and events that

go beyond the information contained within specific modalities. to consist of embodied representations grounded in action and
Amodal representations provide a means of gathering and integrat- perception mechanisms. Glenberg et al. (2008, p. 4) offer the fol-
ing information from different modalities as well as transferring lowing summary of what researchers mean when they say that
information between distinct sensorimotor codes. Because linguis- language is embodied:
tic representations have the design features outlined above, they
Linguistic symbols are embodied to the extent that: (a) the meaning
can also carry out this function (Carruthers, 2002). of the symbol (the interpretant) to the agent depends on activity
I propose that when an individual acquires a natural language, in systems also used for perception, action, and emotion, and (b)
she acquires a representational system that is different in some reasoning about meaning, including combinatorial processes of
important respects from the multimodal, context-sensitive embod- sentence understanding requires use of those systems.
ied symbol systems that exist independently of language. The acqui-
sition of natural language, in other words, enhances and extends The idea is that linguistic symbols have meaning because they
her representational abilities by giving her access to a context-free dynamically activate sensorimotor representations associated with
and arbitrary symbol system. This symbol system is independent interacting with the world. On this account, linguistic symbols are
of, and yet interacts with, other embodied symbols. intermediaries that do not directly have meaning or participate in
This proposal requires a revisionist conception of linguistic com- reasoning about meaning.
petence. Standard theories of linguistic competence are thoroughly I suggest that language plays two roles in our cognitive lives.
amodal. Linguists have identified structural regularities at several One role is to engage sensorimotor simulations of interacting with
levels of analysis, including phonology, morphology, syntax, and to the world. In this role, language serves primarily as a medium of
some degree logical form or semantics. Knowledge relating to these communication. A second role is to elicit and engage symbolically
levels is thought to be contained with language-specific functional mediated associations and inferences. Our concepts are not merely
modules (Fodor, 1983) and is generally thought to be couched couched in sensorimotor representations but also in linguistic
in amodal codes. Comprehension involves translating perceptual representations (words, phrases, sentences). Conceptual content
information into these codes and production involves translating is captured in part by the relationships of linguistic representations
information in these codes into motor representations. The revi- with other linguistic representations. These relationships may be
sionist approach taken in this essay is that the process of achieving merely associative or they may be inferential. On this view, a con-
competence in a specific natural language involves acquiring the cept such as DOG will, not only be represented on a given occasion
ability to generate appropriate simulations of linguistic experience. by multimodal simulations associated with interacting with dogs,
To be successful, these simulations must comport with the struc- but will also be represented in terms of related linguistic words,
tural regularities at the different levels of analysis. They will not, phrases, or sentences. This idea has a clear affinity with inferential
however, depend on knowledge contained with an amodal symbol role or conceptual role semantics (Harman, 1982; Block, 1986). This
system. Three points about this revisionist proposal are especially philosophical theory of mental content holds that the meaning of a
important. The first is that it is neutral with respect to the issue concept is determined by its functional role within the cognitive life
of the degree to which linguistic competence is innate or learned. of an individual. My proposal is distinct from this theory because
This proposal has to do with the format of the representations it adds the further requirement that the associative and inferential
associated with this competence and not how it is acquired. The relationships be couched in language-based simulations.
second is that, despite superficial appearances, this is not an inner One source of evidence for the view that internalized natu-
speech view. The claim is that linguistic competence is contained ral language can itself serve as a symbolic form of cognition is
within a system for generating perceptual symbols. These symbols the effectiveness of statistical models that derive the meaning of
consist of neurophysiological simulations that can be partial, selec- words through statistical computations applied to large corpuses
tive, and unconscious. The third important point is that there is no of text (Louwerse and Jeuniaux, 2008). A prominent example of
independent lexical semantic code. The core thesis of this paper is this type of model is Latent Semantic Analysis or LSA (Landauer
that concepts are couched in two types of simulation-based repre- and Dumais, 1997). The idea behind LSA is that the aggregate of
sentations: those associated with non-linguistic experience of the all the linguistic contexts in which a given word does and does
world and those associated with experience of language. Because not appear constrains semantic-relatedness. LSA has shown some
simulations are detailed and often complex, linguistic perceptual effectiveness with respect to modeling a variety of linguistic tasks
symbols may exhibit structure at the various levels of analysis (pho- (Landauer et al., 1998). For example, an LSA model performed at
nology, morphology, syntax, etc.). a comparable level on the vocabulary portion of the Test of English
as a Foreign Language to a large sample of students applying for
THINKING IN WORDS college entrance in the United States from non-English speaking
Despite the clear differences between embodied and orthodox countries (Landauer and Dumais, 1997). Even if we grant that this
approaches to cognition, both adopt a similar view of the relation- particular model is psychologically implausible, it demonstrates the
ship between language and thought. Both see language as a medium potential of a language-based representational system.
of communication rather than a medium of thought. According to
both, language expresses underlying thoughts that are encoded in THEORETICAL INFLUENCES
some other semantic code. Within traditional cognitive science, this I propose that our concepts are encoded in at least two types of
code is typically taken to be a language-like amodal symbol system semantic representations: one type employing embodied sensori-
(Fodor, 1975). Within embodied cognition, this code is thought motor representations associated with our experience of the world

and the other type employing dis-embodied sensorimotor repre- causal relationship with objects and events. Clark (2008) argues
sentations associated with our experience of language. Other types that language helps extend our cognitive abilities in at least three
may exist. Gesture, for instance, might form an independent seman- distinct but related ways: first, the very act of labeling objects and
tic representational system (Goldin-Meadow, 2003). This pluralistic events provides a means of discovering increasingly abstract pat-
embodied proposal has clear similarities with some previous theo- terns in nature; second, the ability to recall and react to structured
ries. Highlighting the similarities – and the differences – between it sentences enables us to acquire new skills and capacities, and third,
and these theories should help clarify its central claims. our language abilities partially underwrite our ability to reflect
This proposal overlaps somewhat with another recent attempt to on and influence the contents of our own thinking. Because he
offer an embodied solution to the problem of abstraction. Borghi is primarily interested in simply establishing that language can in
and Cimatti (2009) argue that supporters of embodied cognition fact extend our cognitive abilities, Clark focuses on a collection of
have paid too little attention to the embodied social experience empirically based examples that seem to demonstrate cognitive
associated with language. They propose that there is a qualitative extension. One of the most established of these is the apparent way
distinction to be made, not between two different mental processes, in which verbal counting helps children acquire an understanding
but rather between two different cognitive sources of grounding: of positive integers (Dehaene, 1999; Carey and Sarnecka, 2006).
one that depends crucially on direct sensorimotor experience and Where my account diverges from Clark’s is with respect to
another that depends crucially on linguistic experience. Both of scope. I contend that the sort of scaffolding he discusses is not
these sources can be useful in the acquisition of any concept but the limited to specific concepts or cognitive domains. Instead, acquir-
acquisition of concrete concepts is likely to depend more on direct ing a natural language extends our abilities to acquire concepts
sensorimotor experience and the acquisition of abstract concepts across the board. This is not simply because it offers a means of
is more likely to depend on linguistic experience. This distinction accessing socially derived information but also because it offers new
seems important and necessary. I suggest that it falls short, however, representational powers. I suggest that most concepts depend to
because it does not appropriately emphasize the importance of some significant degree on information represented in internalized
the computational properties of natural language. While I agree natural language.
that linguistic experience is an important source of socially derived Clark may or may not be sympathetic with this general point, but
information about the world, I maintain that the structural proper- there is no indication that he connects this scaffolding effect to the
ties of natural language contribute to its effectiveness in represent- qualitative distinction between abstract and concrete concepts.
ing abstract concepts. My account differs from Borghi and Cimatti’s This brings us to perhaps the single greatest influence of the
because it holds that the acquisition of language creates a new theory outlined in this essay: Dual Coding Theory or DCT (Paivio,
dis-embodied semantic system, one that has many of the proper- 1986). This theory posits two independent cognitive subsystems,
ties usually associated with the amodal symbol systems favored by one employing symbolic verbal representations and the other
traditional cognitive science. In other words, natural language on employing analog non-verbal representations. Sadoski and Paivio
my view is not merely another source of information about the (2004, p. 1340) write:
world but is also another way of thinking about the world.
A basic premise of DCT is that all mental representations retain some
My core thesis is that language is an internalized amodal symbol of the concrete qualities of the external experiences from which they
system that is built on an embodied substrate. As such, it extends derive. These experiences can be linguistic or non-linguistic. Their
our cognitive reach and helps us overcome the problem of abstrac- different characteristics develop into two separate mental systems,
tion. This idea is inspired in part by Andy Clark’s view of language or codes, one specialized for representing and processing language
as a kind of cognitive scaffolding that provides cognitive benefits (the verbal code) and one for processing non-linguistic objects and
that would not otherwise be available to us. Clark (2008, p. 47) events (the non-verbal code).
summarizes these benefits in the following passage:
The focus in DCT on the dynamic relationship between experi-
The computational value of a public system of essentially context-
free, arbitrary symbols, lies{ in the way such a system can push, ence and mental representations seems to be in keeping with the
pull, tweak, cajole, and eventually cooperate with various non- basic tenets of embodied cognition. One might even reasonably
arbitrary, modality-rich, context-sensitive forms of biologically see DCT as a precursor to the embodied cognition movement.
basic encoding. However, an important aspect of DCT, i.e., its emphasis on lan-
guage as an independent symbol system, has not generally been
Clark’s claim is that natural language augments the cognitive abili- taken up by embodied cognition. To a certain degree, my account
ties of an embodied mind. The core idea is that natural language can be seen as an attempt to recapture an important insight from
is a cognitively useful symbol system, not because it mirrors the DCT within an embodied framework. It is important, however, to
structure of our underlying thoughts, but because it does not. Clark recognize that the result of this effort is not simply a recapitula-
makes much of the arbitrariness of linguistic symbols. Although the tion of DCT. There are some important differences between the
arbitrariness of the relationship between words and their seman- account developed here and DCT. First, DCT claims that mental
tic contents is well known, one might think that “forms of bio- images are the basic constituents of the verbal and non-verbal
logically basic encoding” are equally arbitrary. However, as we saw systems. My account views perceptual symbols as the basic units.
above in the context of the symbol grounding problem, there is a This is significant because perceptual symbol system theory rep-
sense in which perceptual symbols are not arbitrary because they resents an explicit attempt to avoid the weaknesses associated with
contain sensorimotor representations that enjoy a non-cognitive image-based theories of concepts. Perceptual symbols differ from

mental images in a number of important ways: for instance, they grammatical deficits are associated with damage throughout
need not be conscious, they can be schematic, and they are often the left perisylvian cortex (Caplan et al., 1996). Finally, recent
multi-modal. Second, DCT and my theory differ with respect to evidence suggests that Broca’s area itself might have multiple
the nature of the mental representations associated with language. functions. For example, a number have studies have implicated
According to DCT, they are a special class of mental images that in action-related tasks (Thoenissen et al., 2002; Nishitani et al.,
are made up from different basic elements (logogens) than the 2005). In sum, evidence from cognitive neuroscience and neu-
basic elements of non-verbal representations (imagens). On my ropsychology suggests that language processing is widely dis-
account, all conceptual representations consist of perceptual tributed in the brain and involves a number of sensorimotor
symbols. Linguistic representations are distinguished form non- areas. Although this distribution is not logically incompatible
linguistic ones by the fact that they are an internalization of an with an amodal approach, it fits well with the idea that language
external symbol system. processing involves sensorimotor simulations.
In the end, the view advocated in this essay brings together ideas A second, more direct reason to think that language process-
from a number of different theories and combines them in a novel ing might involve perceptual symbols is that there is evidence
way. While it clearly owes a debt to these previous views, it stands of functional links between motor and perception circuits with
or falls on its own. the left perisylvian cortex (Pulvermüller, 2005). For example,
there is evidence that listening to speech modulates tongue
EMPIRICAL EVIDENCE muscle responses (Fadiga et al., 2002). This sort of evidence is
We began the last section with the acknowledgment of the seri- often seen as supporting the motor theory of speech percep-
ousness of the problem of abstraction. We now have a theoretical tion (Liberman and Whalen, 2000) or the direct realist theory
picture of how language might help explain this ability: language (Fowler et al., 2003). Critics of these theories argue that auditory
might extend our cognitive abilities in such a way that enables us areas alone might be sufficient for perceiving speech (e.g., Toni
to have some of the benefits of an amodal symbol system. This et al., 2008). If true, this would rule out a strongly action-based
theoretical picture rests on two independent hypotheses: (1) that account of speech perception in which speech perception nec-
language processing involves sensorimotor simulation and (2) that essarily involves motor processing. However, it does not rule
linguistic representations play an important role in our ability to out a weaker view that speech recognition generally involves
abstract and generalize. multimodal perceptual symbols.
A third reason to suppose that language processing involves per-
LANGUAGE PROCESSING INVOLVES PERCEPTUAL SYMBOLS ceptual symbols is that several studies implicate active integration
Given the dynamic nature of linguistic communication, the idea of multimodal information in on-line language processing. It is
that language processing involves perceptual symbols seems attrac- well established that visual input can influence phonemic speech-
tive. After all, most linguistic communication is time-constrained processing (McGurk and MacDonald, 1976). A large body of eye-
and would seem to require the integration of action, perception, tracking experiments shows the manifold ways in which visual
and cognition. Below, I survey some of the evidence favoring information can be continuously integrated with auditory infor-
this hypothesis. mation during the processing of speech (Spivey and Richardson,
The first reason to think language processing might involve 2009). Visual information has been shown to influence language
sensorimotor simulations is a negative one: the project to locate comprehension at various levels of linguistic analysis, including
self-contained language areas of the brain has not succeeded. word-recognition (Allopenna et al., 1998), syntactic processing
Ever since the work of Broca and Wernicke in the late nineteenth (Tanenhaus et al., 1995), and thematic role assignment (Altmann
century (Finger, 1994), the classical localizationist position has and Kamide, 1999). Consider a study involving syntactic ambigu-
been that subcomponents of language are represented and proc- ity (Spivey et al., 2002; Spivey and Richardson, 2009). Participants
essed in bounded and specialized cortical regions (Geschwind, were presented with a four-quadrant display of real objects and
1970). One of the primary sources of evidence for this perspec- instructed to carry out actions. The display on one condition con-
tive has been the study of aphasic syndromes resulting from tained (going clockwise from the upper left quadrant) a spoon
focal brain injuries (for a review see Saffran, 2000). Researchers, on a napkin, a bare napkin, a bowl, and a pen. The participants
however, have begun to move away from strict localization and were instructed to “Put the spoon on the napkin in the bowl.”
toward the view that language requires the activity of a number Eye-tracking evidence indicates that subjects often fixate on the
of spatially distinct brain regions. This shift has occurred in irrelevant bare napkin before fixating on the bowl and carrying
response to several forms of evidence. For one, neuroimaging out the action. This suggests that they initially misparse the initial
studies indicate that widely distributed brain areas are active prepositional phrase as syntactically attached to the verb. This effect
in language processing (Posner and Raichle, 1994). Another did not occur with a similar display in which two spoons appear,
reason for this shift is the fact that the association of gram- one on a napkin and one not on a napkin (replacing the pen in
matical processing with Broca’s area has broken down to a large the earlier display).
degree (Grodzinsky, 2000). For instance, there is evidence of A fourth reason to think that language processing might involve
some retained grammatical knowledge in Broca’s aphasics (Bates perceptual symbols is the employment of perceptual areas in lan-
and Wulfeck, 1989; Bates et al., 1991). In addition, grammati- guage processing among people with congenital perceptual defi-
cal deficits have been found in Wernicke’s aphasics and other cits. For example, neuroimaging studies find increased activation
clinical populations (Dick et al., 2001). It also appears that in auditory areas when congenitally deaf individuals view signs

(Petito et al., 2000). Similarly, some primary visual areas show highly imageable words have greater contextual information stored
increased activation when congenitally blind individuals read in semantic memory, and imageability effects are to be explained by
Braille (Sadato et al., 1996). the facilitation of processing associated with increased activation
Taken together, these various bodies of evidence suggest that in these networks. On this approach, the reason that participants
language processing is much more integrated with action and respond more quickly in a lexical decision task to a word such as
perception systems than was previously assumed by research- “fingertip” than to one such as “idea” is that the former has more
ers. It should be acknowledged, however, that this evidence is semantic associations than the latter.
only suggestive and not conclusive. One could maintain that this Evidence suggests that both theories are right, depending on the
evidence does not falsify the hypothesis that language process- task. I am going to focus on the evidence for the DCT because this
ing is handled by amodal symbols since the implicated activ- evidence has more relevance to the claims in this essay.
ity in sensorimotor systems could be associated with spreading Consider first neuropsychological case studies. Several research
activation and not be constitutive of language processing. As I teams describe aphasic patients with significant left hemisphere
mentioned earlier in the essay, this is a general problem faced by damage who exhibit a selective semantic impairment for high
any embodied hypothesis. Ultimately, the issue is an empirical imageable words (Berndt et al., 2002; Bird et al., 2003; Crepaldi
one, and unfortunately the evidence currently available does not et al., 2006). Patients with a selective semantic impairment for
completely settle matters. low imageable words are less common but have also been found
Given this uncertainty, it seems worthwhile to consider what (Marshall et al., 1996; Luzzatti et al., 2002). This double dissocia-
would happen if it turns out that language processing is indeed tion suggests that, at least at some level, the semantic processing of
handled by an amodal symbol system of the sort posited by the concepts with low imageability is functionally independent from
current orthodoxy. This would turn the hypothesis that language is the semantic processing of concepts with high imageability.
a form of dis-embodied cognition into the hypothesis that language A number of event-related potential (ERP) experiments sup-
is a form of disembodied cognition (non-hyphenated). It would port a neuroanatomical distinction between concepts of high and
result in a different kind of hybrid theory, one in which concepts are low imageability. For instance, Holcomb et al. (1999) created a task
represented by both multimodal perceptual symbols and amodal that involved manipulations of both context and concreteness.
linguistic symbols. Although I am promoting the dis-embodied ERP recordings were time-locked to sentence-final words in a
view in this essay, the second view is an intriguing and compelling word-by-word reading task in which participants made semantic
alternative (for general arguments in favor of a hybrid approach congruency judgments (e.g., Armed robbery implies that the thief
see Dove, 2009; Kemmerer, 2010). used a weapon vs. armed robbery implies that the thief used a
rose). They found that sentence-final concrete words generated
IMAGEABILITY RECONSIDERED a larger and more anterior N400 than sentence-final abstract
Imageability effects provide support for the account developed in words in both contexts (see also Kounios and Holcomb, 1994;
this essay. Typically, imageability is defined as the ease with which West and Holcomb, 2000). Further studies have found context-
a word gives rise to a sensory-motor mental image (Paivio, 1971). independent topographic effects associated with imageability in
Imageability is a broader concept than concreteness because it single-word presentations (Kellenbach et al., 2002; Swaab et al.,
includes sensory images of bodily states and motor images. It 2002). Using two-word stimuli that involved a noun preceded by
is generally recognized that imageability better captures the rel- either a concrete modifier or an abstract modifier (“green book”
evant phenomena and supports broader generalizations. Highly vs. “engaging book”) in a visual half-field presentation, Huang
reliable imageability ratings on number scales have been gathered et al. (2010) found distinct hemispheric responses. Thus, ERP
for linguistic concepts by number of researchers (Toglia and studies employing diverse tasks support the notion that different
Battig, 1978; Bird et al., 2001). Traditionally, cognitive scientists cognitive systems are associated with the semantic processing of
examined imageability in terms of processing advantages for high high and low imageable words.
imageable concepts over low imageable ones in several cognitive Neuroimaging data supports the notion that neural activity is
tasks. For instance, lexical access has been shown to be quicker for modulated by imageability. A number of studies find that abstract
highly imageable words than for abstract ones (Coltheart et al., or low imageable words elicit greater activation than concrete
1980) and highly imageable words are recalled more quickly in or high imageable words in superior regions of the left tempo-
memory tasks than abstract words (Paivio, 1986; Wattenmaker ral lobe (Mellet et al., 1998; Giesbrecht et al., 2004; Noppeney
and Shoben, 1987). and Price, 2004; Binder et al., 2005; Sabsevitz et al., 2005) and
Two major theories dominate the literature: the DCT (Paivio, inferior regions of the left prefrontal cortex (Giesbrecht et al.,
1971, 1986) and the context-availability theory (Schwanenflugel 2004; Noppeney and Price, 2004; Binder et al., 2005; Sabsevitz
and Shoben, 1983). According to the DCT, words with low image- et al., 2005; Goldberg et al., 2006). This evidence fits with imag-
ability are associated primarily with verbal representations while ing studies that implicate the left inferior frontal gyrus or IFG in
highly imageable words are associated with both verbal representa- language processing (Bookheimer, 2002). When researchers make
tions and perceptual ones. Imageability effects are then explained in the comparison in the reverse direction, the pattern is less clear.
terms of the greater availability of perceptually encoded informa- Whereas some studies find no areas of increased activation (Kiehl
tion. According to the context-availability theory, highly imageable et al., 1999; Perani et al., 1999; Tyler et al., 2001; Grossman et al.,
words are more closely linked to relevant contextual knowledge in 2002; Noppeney and Price, 2004), others find increased activation
semantic networks than less imageable concepts. In other words, in right hemisphere areas (Mellet et al., 1998; Jessen et al., 2000;

Binder et al., 2005; Sabsevitz et al., 2005). This divergence with CONCLUSION
respect to activation patterns fits with the neuropsychological In this essay, I have attempted to assess the generality of embodied
observation that patients are more likely to have a selective deficit cognition. The current evidence for conceptual embodiment is
for abstract or low imageable words than for concrete or high compelling but, unfortunately, circumscribed. Part of the prob-
imageable words. lem is that there has not been enough research on abstract con-
Sabsevitz et al. (2005) carried out a particularly careful fMRI cepts. Beyond this evidential lacuna, though, abstract concepts
study. Their study incorporated a larger sample (28 adults) than represent an important theoretical challenge to embodied cog-
previous studies and a task (judgment of semantic similarity) nition. The most promising attempts to deal with this problem
that is more likely to elicit deep semantic processing than a more appeal to what I have called dis-embodied representations. I have
superficial task, such as lexical decision. Participants were visually argued that there are good reasons to think that natural language
presented with three words (e.g., cheetah, wolf, and tiger) in the itself is a form of dis-embodied cognition. The acquisition of
form of a triangle. The task was to decide which of the two bot- competence with respect to a natural language provides access
tom words was most semantically similar to the top word. In this to syntactically recombinable symbol system that extends our
task, abstract nouns elicited greater activation in the left superior cognitive reach.
temporal and left inferior frontal cortex than concrete nouns, while The speculation that natural language extends the cognitive capaci-
concrete nouns elicited greater activation in a bilateral network of ties of embodied minds points the way to new research opportunities.
association areas than abstract nouns. One question that needs to be answered more fully is just how the
The upshot of this survey is that imageability effects have two types of conceptual symbol systems interact. The potential for
been found in multiple disciplines by investigators in a number interaction is implicit in the dual functionality of linguistic symbols.
of labs using different research methodologies and measures. On the account developed here linguistic representations can serve as
These effects provide support for the notion that abstract or low elicitors of non-linguistic perceptual symbols and as semantic symbols
imageability concepts are processed somewhat differently than in their own right. Presumably, we have the ability to employ these
concrete or high imageability concepts. Areas associated with systems in a context-sensitive and flexible way. However, the nature of
language processing appear to be more active during semantic this flexibility remains to be seen. Another question that arises is the
tasks associated with abstract or low imageability concepts. This extent to which language might explain other significant features of
pattern of activation fits with both the hypothesis that language cognition. For example, both Dennett (1996) and Carruthers (2002)
is a dis-embodied form of cognition and the hypothesis that it suggest that language may be the medium for conscious deliberation.
is an amodal form of cognition. The decision between these two Although this is not implied by the position outlined in this essay,
hypotheses turns on the role played by the observed activity in the possibility that conscious deliberation involves language-based
language related areas. Is it part of linguistic sensorimotor simula- perceptual symbols seems worthy of investigation. In the end, the
tions or is it part of amodal linguistic processes? This question hypothesis that language is a dis-embodied form of cognition has
awaits further research. both empirical support and theoretical promise.
REFERENCES implications for category structure. in modality-specific systems. Trends Bird, H., Franklin, S., and Howard, D.
Adolphs, R., Damasio, H., Tranel, D., Mem. Cogn. 10, 82–93. Cogn. Sci. 7, 84–91. (2001). Age of acquisition and image-
Cooper, G., and Damasio, A. R. Barsalou, L. W. (1999). Perceptual sym- Barsalou, L. W., and Wiemer-Hastings, K. ability ratings for a large set of words,
(2000). A role for somatosensory bol systems. Behav. Brain Sci. 22, (2005). “Situating abstract concepts,” including verbs and function words.
cortices in the visual recognition of 577–660. in Grounding Cognition: The Role of Behav. Res. Methods 33, 73–79.
emotion as revealed by three-dimen- Barsalou, L. W. (2003). Abstraction in Perception and Action in Memory, Bird, H., Howard, D., and Franklin,
sional lesion mapping. J. Neurosci. 20, perceptual symbol systems. Philos. Language, and Thought, eds D. Pecher S. (2003). Verbs and nouns: the
2683–2690. Trans. R. Soc. Lond., B, Biol. Sci. 358, and R. Zwaan (Cambridge, UK: importance of being imageable. J.
Allopenna, P. D., Magnuson, J. S., and 1177–1187. Cambridge University Press), 129–163. Neurolinguistics 16, 113–149.
Tanenhaus, M. K. (1998). Tracking the Barsalou, L. W., and Prinz, J. J. (1997). Bates, E., and Wulfeck, B. (1989). Block, N. (1986). Advertisement for a
time course of spoken word recogni- “Mundane creativity in perceptual Comparative aphasiology: a cross- semantics for psychology. Midwest
tion using eye movements: evidence symbol systems,” in Conceptual linguistic approach to language break- Stud. Philos. 10, 615–78.
for continuous mapping models. J. Structures and Processes: Emergence, down. Aphasiology 3, 111–142. Bookheimer, S. (2002). Functional MRI of
Mem. Lang. 38, 419–439. Discovery, and Change, eds T. B. Ward, Bates, E., Wulfeck, B., and MacWhinney, language: new approaches to under-
Altmann, G. T. M., and Kamide, Y. (1999). S. M. Smith, and J. Vaid (Washington, B. (1991). Cross-linguistic studies of standing the cortical organization
Incremental interpretation at verbs: DC: American Psychological aphasia: an overview. Brain Lang. 41, of semantic processing. Annu. Rev.
restricting the domain of subsequent Association), 267–307. 123–148. Neurosci. 25, 151–188.
reference. Cognition 73, 247–264. Barsalou, L. W., Santos, A, Simmons, Berndt, R. S., Haendiges, A. N., Burton, Borghi, A. M. (2005). “Object concepts
Bak, T. H., O’Donovan, D. G., Xuereb, W. K., and Wilson, C. D. (2008). M. W., and Mitchum, C. C. (2002). and action,” in Grounding Cognition:
J. H., Boniface, S., and Hodges, J. R. “Language and simulation in con- Grammatical class and imageability in The Role of Perception and Action
(2001). Selective impairment of verb ceptual processing,” in Symbols, aphasic word production: their effects in Memory, eds D. Pecher and R. A.
processing associated with pathologi- Embodiment, and Meaning, eds M. are independent. J. Neurolinguistics 15, Zwaan (Cambridge, UK: Cambridge
cal changes in Brodmann areas 44 De Vega, A. M. Glenberg, and A. 353–371. University Press), 8–34.
and 45 in the motor neurone disease- C. Graesser (Oxford, UK: Oxford Binder, J., Westbury, C., McKiernan, K., Borghi, A. M., and Cimatti, F. (2009).
dementia-aphasia syndrome. Brain University Press), 245–284. Possing, E., and Medler, D. (2005). “Words as tools and the prob-
124, 103–120. Barsalou, L. W., Simmons, W. K., Barbey, Distinct brain systems for process- lem of abstract words meanings,”
Barr, R. A., and Caplan, L. J. (1987). A. K., and Wilson, C. D. (2003). ing concrete and abstract concepts. J. in Proceedings of the 31st Annual
Category representations and their Grounding conceptual knowledge Cogn. Neurosci. 17, 905–917. Conference of the Cognitive Science

Society, eds N. Taatgen and H. van Crepaldi, D., Aggujaro, S., Arduino, L. Geschwind, N. (1970). The organization and abstract concepts. Neuroimage 49,
Rijn (Amsterdam: Cognitive Science S., Zonca, G., Ghirardi, G., Inzaghi, of language and the brain. Science 170, 1116–1123.
Society), 2304–2309. M. G., Colombo, M., Chierchia, G., 940–944. Jackendoff, R. (1992). Languages of the
Borghi, A. M., Glenberg, A. M., and and Luzatti, C. (2006). Noun–verb Giesbrecht, B., Gamblin, C., and Swaab, T. Mind: Essays on Mental Representation.
Kaschak (2004). Putting words in per- dissociation in aphasia: The role of (2004). Separable effects of semantic Cambridge, MA: MIT Press.
spective. Mem. Cogn. 32, 863–873. imageability and functional location priming and imageability on word Jeannerod, M. (1995). Mental imagery in
Boroditsky, L., and Ramscar, M. (2002). of the lesion. Neuropsychologia 44, processing in human cortex. Cereb. the motor context. Neuropsychologia
The roles of body and mind in abstract 73–89. Cortex 14, 521–529. 33, 1419–1432.
thought. Psychol. Sci. 13, 185–189. Crutch, S. J., and Warrington, E. K. Glenberg, A. M., de Vega, M., and Graesser, Jessen, F., Heun, R., Erb, M., Granath,
Buccino, G., Riggio, L., Melli, G., Binkofski, (2005). Abstract and concrete con- A. C. (2008). “Framing the debate,” in D.-O., Klose, U., Papassotiropoulos,
F., Gallese, V., and Rizzolatti, G. (2005). cepts have structurally different rep- Symbols, Embodiment, and Meaning, A., and Grodd, W. (2000). The con-
Listening to action-related sentences resentational frameworks. Brain 128, eds M. De Vega, A. M. Glenberg, and crétense effect: evidence for dual
modulates the activity of the motor sys- 615–627. A. C. Graesser (Oxford, UK: Oxford coding and context availability. Brain
tem: a combined TMS and behavioral Damasio, A. R., and Damasio, H. (1994). University Press), 1–10. Lang. 74, 103–112.
study. Cogn. Brain Res. 24, 355–363. “Cortical systems for retrieval of con- Glenberg, A. M., and Kaschak, M. P. Kiehl, K. A., Liddle, P. F., Smith, A. M.,
Camp, E. (2009). Putting thoughts to crete knowledge: the convergence zone (2002). Grounding language in action. Mendrek, A., Forster, B. B., and Hare,
work: concepts, systematicity, and framework,” in Large-Scale Neuronal Psychon. Bull. Rev. 9, 558–565. R. D. (1999). Neural Pathways involved
stimulus-independence. Philos. Theories of the Brain. Computational Glenberg, A. M., and Robertson, D. A. in the processing of concrete and
Phenomenol. Res. 78, 275–311. neuroscience, eds C. Koch and J. L. (1999). Indexical understanding of abstract words. Hum. Brain Mapp. 7,
Caplan, D., Hildebrandt, N., and Makris Davis (Cambridge, MA: MIT Press), instructions. Discourse Process. 28, 225–233.
N. (1996). Location of lesions in stroke 61–74. 1–26. Kellenbach, M. L., Wijers, A. A., Hovis,
patients with deficits in syntactic Dehaene, S., Spelke, E., Pinel, P., Stanescu, Glenberg, A. M., Sato, M., Cattaneo, L., M., Mulder, J., and Mulder, G. (2002).
processing in sentence comprehen- R., and Tviskin, S. (1999). Sources of Riggio, L., Palumbo, D., and Buccino, Neural differentiation of lexico-syn-
sion. Brain 119, 933–949. mathematical thinking: behavioral G. (2008). Processing abstract language tactic categories or semantic features?
Caramazza, A., and Mahon, B. Z. (2006). and brain-imaging evidence. Science modulates motor system activity. Q. J. Event related potential evidence for
The organization of conceptual 284, 970–974. Exp. Psychol. 61, 905–919. both. J. Cogn. Neurosci. 14, 561–577.
knowledge in the brain: the future’s Deñabeitia, J. A., Aviles, A., Afonso, O., Goldberg, R. F., Perfetti, C. A., and Kemmerer, D. (2010). “How words cap-
past and some future directions. Cogn. Scheepers, C., and Carreiras, M. Schneider, W. (2006). Distinct and ture visual experience: the perspective
Neuropsychol. 23, 13–38. (2009). Qualitative differences in common cortical activations for from cognitive neuroscience,” in Words
Carey, S., and Sarnecka, B. W. (2006). “The the representation of abstract versus multimodal semantic categories. and the World: How Words Capture
development of human conceptual rep- concrete words: evidence from the Cogn. Affect. Behav. Neurosci. 6, Human Experience, eds B. Malt and
resentations: a case study,” in Processes visual-word paradigm. Cognition 214–222. P. Wolff (Oxford: Oxford University
of Change in Brain and Cognitive 110, 284–292. Goldin-Meadow, S. (ed.). (2003). Hearing Press), 287–327.
Development, eds Y. Munakata and Dennett, D. (1996). Kinds of Minds: Toward Gesture: How Our Hands Help Us Think. Kosslyn, S. M. (1994). Image and Brain:
M. Johnson (Oxford, UK: Oxford an Understanding of Consciousness. Cambridge, MA: Belknap Press. The Resolution of the Imagery Debate.
University Press), 473–496. New York: Harper Collins. Grèzes, J., and Decety, J. (2001). Functional Cambridge, MA: MIT Press.
Carruthers, P. (2002). The cognitive func- Dick, F., Bates, E. Wulfeck, B., Utman, J. A., anatomy of execution, mental simula- Kosslyn, S. M., and Shwartz, S. P. (1977).
tions of language. Behav. Brain Sci. 25, Dronkers, N., and Gernsbacher, M. A. tion, observation, and verb generation A simulation of visual imagery. Cogn.
657–674. (2001). Language deficits, localization, of actions: a meta-analysis. Hum. Brain Sci. 1, 265–295.
Casasanto, D., and Boroditsky, L. (2008). and grammar: evidence for a distribu- Mapp. 12, 1–19. Kounios, J., and Holcomb, P. J. (1994).
Time in the mind: using space to think tive model of language breakdown in Grodzinsky, Y. (2000). The neurology of Concreteness effects in semantic
about time. Cognition 106, 579–593. aphasic patients and neurologically syntax: language use without Broca’s processing: ERP evidence supporting
Chao, L. L., and Martin, A. (2000). intact individuals. Psychol. Rev. 108, area. Behav. Brain Sci. 23, 1–71. dual-coding theory. J. Exp. Psychol.
Representation of manipulable man- 759–788. Grossman, M., Koenig, P., DeVita, C., Learn. Mem. Cogn. 20, 804–823.
made objects in the dorsal stream. Dove, G. O. (2009). Beyond perceptual Glosser, G., Alsop, D., Detre, J., and Lakoff, G. (1987). Women, Fire, and
Neuroimage 12, 478–484. symbols: a call for representational Gee, J. (2002). The neural basis for Dangerous Things: What Categories
Chatterjee, A. (2010). Disembodying cog- pluralism. Cognition 110, 412–431. category-specific knowledge: An fMRI Reveal About the Mind. Chicago:
nition. Lang. Cogn. 2, 79–116. Fadiga, L., Craighero, L., Buccino, G., and study. Neuroimage 15, 936–948. University of Chicago Press.
Chemero, A. (2009). Radical Embodied Rizzolatti, G. (2002). Speech listening Harman, G. (1982). Conceptual role Lakoff, G., and Johnson, M. (1980).
Cognitive Science. Cambridge, MA: specifically modulates the excitability semantics. Notre Dame J. Formal Logic Metaphors We Live By. Chicago:
MIT Press. of tongue muscles: a TMS study. Eur. 28, 252–256. University of Chicago Press.
Chomsky, N. (1966). Cartesian Linguistics: J. Neurosci. 15, 399–402. Harnad, S. (1990). The symbol grounding Landauer, T. K., and Dumais, S. T. (1997).
A Chapter in the History of Rationalist Finger, S. (1994). Origins of Neuroscience. problem. Physica D 42, 335–346. A solution to Plato’s problem: the
Thought. New York: Harper and Row. Oxford: Oxford University Press. Hauk, O., Johnsrude, I. and Pulvermüller, Latent Semantic Analysis theory of
Clark, A. (1997). Being There: Putting Fodor, J. A. (1975). The Language of F. (2004). Somatotopic representation acquisition, induction and represen-
Brain, Body, and World Together Again. Thought. Cambridge, MA: Harvard of action words in human motor tation of knowledge. Psychol. Rev. 104,
Cambridge, MA: MIT Press. University Press. and premotor cortex. Neuron 41, 211–240.
Clark, A. (1998). “Embodied, situated, Fodor, J. A. (1983). The Modularity of 301–307. Landauer, T. K., Foltz, P. W., and Laham,
and distributed cognition,” in A Mind: An Essay on Faculty Psychology. Holcomb, P. J., Kounios, J., Anderson, J. E., D. (1998). Introduction to Latent
Companion to Cognitive Science, eds W. Cambridge, MA: MIT Press. and West, W. C. (1999). Dual-coding, Semantic Analysis. Discourse Process
Bechtel and G. Graham (Malden, MA: Fodor, J. A., and Pylyshyn, Z. W. (1988). context- availability, and concreteness 25, 259–284.
Blackwell Publishers), 506–517. Connectionism and cognitive archi- effects in sentence comprehension: An Liberman, A. M., and Whalen, D. H. (2000).
Clark, A. (2008). Supersizing the Mind: tecture: a critical analysis. Cognition electrophysiological investigation. J. On the relation of speech to language.
Embodiment, Action, and Cognitive 28, 3–71. Exp. Psychol. Learn. Mem. Cogn. 25, Trenes Cogn. Sci. 4, 187–196.
Extension. Oxford, UK: Oxford Fowler, C. A., Brown, J. M., Sabadini, L., 721–742. Louwerse, M., and Jeuniaux, P. (2008).
University Press. and Weihing, J. (2003). Rapid access to Huang, H., Lee, C., and Federmeier (2010). “Language comprehension is both
Coltheart, M., Patterson, K., and Marshall, speech gestures in perception: evidence Imagine that! ERPs provide evidence embodied and symbolic,” in Symbols,
J. C. (eds). (1980). Deep Dyslexia. from choice and simple response time for distinct hemispheric contribu- Embodiment, and Meaning, eds M.
London: Routledge and Kegan Paul. tasks. J. Mem. Lang. 4, 396–410. tions to the processing of concrete De Vega, A. M. Glenberg, and A.

C. Graesser (Oxford, UK: Oxford Pecher, D., Zeelenberg, R., and Barsalou, and concrete verbal materials. J. Thoenissen, D., Zilles, K., and Toni, I.
University Press), 309–326. L. W. (2003). Verifying properties from Exp. Psychol. Learn. Mem. Cogn. 9, (2002). Differential involvement
Luzzatti, C., Raggi, R., Zonca, G., Pistarini, different modalities for concepts pro- 82–102. of parietal and precentral regions
C., Contardi, A., and Pinna, G. (2002). duces switching costs. Psychol. Sci. 14, Scorolli, C. (2009). Language and in movement preparation and
Verb–noun double dissociation in 119–124. Embodiment: Sensory-Motor and motor intention. J. Neurosci. 22,
aphasic lexical impairments: the role Perani, D., Cappa, S., Schnur, T., Linguistic-Social Experience. Evidence 9024–9034.
of word frequency and imageability. Tettamanti, M., Collina, S., Rosa, M., on Sentence Comprehension (Doctoral Toglia, M. P., and Battig, W. F. (1978).
Brain Lang. 81, 432–444. and Fazio, F. (1999). The neural cor- dissertation). Available at: http:// Handbook of Semantic Word Norms.
Machery, E. (2007). Concept empiricism: relates of verb and noun processing: a en.scientificcommons.org/55511932 New York: Erlbaum.
a methodological critique. Cognition PET study. Brain 122, 2337–2344. Scorolli, C., and Borghi, A. M. (2007). Toni, I., de Lange, F. P., Noordzij, M., and
104, 19–46. Petito, L.A., Zatorre, R. J., Gauna, K., Nikelski, Sentence comprehension and action: Hagoort, P. (2008). Language beyond
Machery, E. (2010). Précis of “doing E. J., Dostie, D., and Evans, A. C. (2000). effector specific modulation of action. J. Physiol. Paris 102, 71–79.
without concepts.” Behav. Brain Sci. Speech-like cerebral activity in pro- the motor system. Brain Res. 1130, Tyler, L. K., Russell, R., Fadili, J., and Moss,
33, 195–244. foundly deaf people processing signed 119–124. H. (2001). The neural representation
Maguire, M. J., and Dove, G. O. (2008). languages: implications for the neural Scorolli, C., Borghi, A. M., and Glenberg, of nouns and verbs: PET studies.
“Speaking of events: event word basis of human language. Proc. Natl. A. M. (2009). Language-induced Brain, 124, 1619–1634.
learning and event representation,” in Acad. Sci. U.S.A. 97, 13961–13966. motor activity in bimanual object lift- van Dantzig, S., Pecher, D., Zeelenberg, R.,
Understanding Events: How Humans Pezzulo, G., and Castelfranchi, C. (2007). ing. Exp. Brain Res. 193, 43–53. and Barsalou, L. W. (2008). Perceptual
See, Represent, and Act on Events, eds The symbol detachment problem. Simmons, W. K., and Barsalou, L. W. processing affects conceptual process-
T. F. Shipley and J. Zacks (Oxford, UK: Cogn. Process 8, 115–131. (2003). The similarity-in-topography ing. Cogn. Sci. 32, 579–590.
Oxford University Press), 193–218. Pinker, S. (1994). Language Instinct. New principle: reconciling theories of con- Wattenmaker, W., and Shoben, E. (1987).
Mahon, B. Z., and Caramazza, A. (2008). York: Harper Collins. ceptual deficits. Cogn. Neuropsychol. Context and the recallability of con-
A critical look at the embodied cogni- Posner, M. I., and Raichle, M. E. (1994). 20, 451–486. crete and abstract sentences. J. Exp.
tion hypothesis and a new proposal Images of Mind, 2nd Edn. New York: Simmons, W. K., Ramjee, V., Beauchamp, Psychol. 13, 140–150.
for grounding conceptual content. J. W. H. Freeman and Co. M. S., McRae, K., Martin, A., and Weiskopf, D. (2007). Concept empiri-
Physiol. Paris 102, 59–70. Prinz, J. J. (2002). Furnishing the Mind: Barsalou, L. W. (2007). A common cism and the vehicles of thought. J.
Markman, A., and Dietrich, E. (2000). Concepts and Their Perceptual Basis. neural substrate for perceiving and Conscious. Stud. 14, 156–183.
In defense of representation. Cogn. Cambridge, MA: MIT Press. knowing about color. Neuropsychologia West, W. C., and Holcomb, P. J. (2000).
Psychol. 40, 138–171. Pulvermüller, F (2005). Brain mechanisms 45, 2802–2810. Imaginal, semantic, and surface-level
Marshall, J., Pring, T., Chiat, S., and linking language and action. Nat. Rev. Spivey, M., and Richardson, D. (2009). processing of concrete and abstract
Robson, J. (1996). Calling a salad Neurosci. 6, 576–582. “Language processing embodied words: an electrophysiological
a federation: an investigation of Pulvermüller, F., Hauk, O., Nikulin, V. V., and embedded,” in The Cambridge investigation. J. Cogn. Neurosci. 12,
semantic jargon. Part 1 – nouns. J. Ilmonlemi, R. J. (2005). Functional Handbook of Situated Cognition, eds P. 1024–1037.
Neurolinguistics 9, 237–250. links between motor and language sys- Robbins and M. Aydede (Cambridge, Wiemer-Hastings, K., and Xu, X. (2005).
Martin, A., Wiggs, C. L., Ungerleider, L. tems. Eur. J. Neurosci. 21, 793–797. UK: Cambridge University Press), Content differences for abstract and
G., and Haxby, J. V. (1996). Neural cor- Pylyshyn, Z. W. (1973). What the mind’s 382–400. concrete concepts. Cogn. Sci. 29,
relates of category-specific knowledge. eye tells the mind’s brain: a critique Spivey, M., Tanenhaus, M., Eberhard, K., 719–736.
Nature 379, 649–652. of mental imagery. Psychol. Bull. 80, and Sedivy, J. (2002). Eye-movements Wilson, M. (2002). Six views of embod-
McGurk, H., and MacDonald, T. (1976). 1–24. and spoken language-comprehension: ied cognition. Psychon. Bull. Rev. 9,
Hearing lips and seeing voices. Nature Pylyshyn, Z. W. (1981). The imagery effects of visual context on syntactic 625–636.
264, 746–748. debate: analogue media versus tacit ambiguity resolution. Cogn. Psychol. Winner, E., Rosenstiel, A. K., and Gardner,
Mellet, E., Tzourio, N., Denis, M., and knowledge. Psychol. Rev. 88, 16–45. 45, 447–481. H. (1976). The development of meta-
Mazoyer, B. (1998). Cortical anatomy Richardson, D. C., Spivey, M. J., Barsalou, Stanfield, R. A., and Zwaan, R. A. (2001). phoric understanding. Dev. Psychol.
of mental imagery of concrete nouns L. W., and McRae, K. (2003). Spatial The effect of implied orientation 12, 289–297.
based on their dictionary definition. representations activated during real- derived from verbal context on pic-
Neuroreport 9, 803–808. time comprehension of verbs. Cogn. ture recognition. Psychol. Sci. 12, Conflict of Interest Statement: The
Murphy, G. L. (1996). On metaphoric rep- Sci. 27, 767–780. 153–156. author declares that the research was
resentation. Cognition 60, 173–204. Sabsevitz, D., Medler, D., Seidenberg, M., Swaab, T., Baynes, K., and Knight, R. conducted in the absence of any com-
Murphy, G. L. (1997). Reasons to doubt and Binder, J. (2005). Modulation of (2002). Separable effects of priming mercial or financial relationships that
the present evidence for metaphoric the semantic system by word image- and imageability on word process- could be construed as a potential conflict
representation. Cognition 62, 99–108. ability. Neuroimage 27, 188–200. ing: an ERP study. Cogn. Brain Res. of interest.
Nishitani, N., Schürmann, M., Amunts, Sadato, N., Pascual-Leone, A., Grafman, 15, 99–103.
K., and Hari, R. (2005). Broca’s region: J., Ibanez, V., Deiber, M., Dold, G., and Taddeo, M., and Floridi, L. (2005). The Received: 16 July 2010; accepted: 23
from action to language. Physiology 20, Hallett, M. (1996). Activation of the pri- symbol grounding problem: a critical December 2010; published online: 25
60–69. mary visual cortex by Braille reading in review of fifteen years of research. J. January 2011.
Noppeney, U., and Price, C. J. (2004). blind subjects. Nature 380, 526–528. Exp. Artif. Intell. 17, 419–445. Citation: Dove G (2011) On the need for
Retrieval of abstract semantics. Sadoski, M., and Paivio, A. (2004). “A Tanenhaus, M., Spivey-Knowlton, M., embodied and dis-embodied cognition.
Neuroimage 22, 164–170. dual coding theoretical model of Eberhard, K., and Sedivy, J. (1995). Front. Psychology 1:242. doi: 10.3389/
Paivio, A. (1971). Imagery and Verbal reading,” in Theoretical Models and Integration of visual and linguis- fpsyg.2010.00242
Processes. New York: Holt, Rinehart Processes of Reading, 5th Edn., eds R. B. tic information in spoken lan- This article was submitted to Frontiers
and Winston. Ruddell and N. J. Unrau (Newark, DE: guage comprehension. Science 268, in Cognition, a specialty of Frontiers in
Paivio, A. (1986). Mental Representations: International Reading Association), 1632–1634. Psychology.
A Dual-Coding Approach. New York: 1329–1362. Tettamanti, M., Buccino, G., Saccuman, Copyright © 2011 Dove. This is an open-
Oxford University Press. Saffran, E. M. (2000). Aphasias and the M. C., Gallese, V. Danna, M., Scifo, access article subject to an exclusive license
Pecher, D., Van Dantzig, S., and relationship of language and brain. P., Fazio, F., Rizzolatti, G., Cappa, agreement between the authors and
Schifferstein, H. N. J. (2009). Concepts Semin. Neurol. 20, 409–418. S., and Perani, D. (2005). Listening Frontiers Media SA, which permits unre-
are not represented by conscious Schwanenflugel, P. J., and Shoben, E. to action-related sentences activates stricted use, distribution, and reproduc-
imagery. Psychon. Bull. Rev. 16, (1983). Differential context effects fronto-parietal motor circuits. J. Cogn. tion in any medium, provided the original
914–919. in the comprehension of abstract Neurosci. 17, 273–281. authors and source are credited.
View publication stats

3.carnie Chapter1 Introduction

Uploaded by

Copyright:

Available Formats

3.carnie Chapter1 Introduction

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

3.carnie Chapter1 Introduction

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

On the need for embodied and disembodied cognition

Article in Frontiers in Psychology · January 2011

Book on Abstract Concepts and Grounded Cognition View project

The user has requested enhancement of the downloaded file.

On the need for embodied and dis-embodied cognition

INTRODUCTION clearly demonstrated by the heated nature of the debate concern-

www.frontiersin.org January 2011 | Volume 1 | Article 242 | 1

Frontiers in Psychology | Cognition January 2011 | Volume 1 | Article 242 | 2

EMPIRICAL EVIDENCE Using a different experimental paradigm, Buccino et al. (2005)

www.frontiersin.org January 2011 | Volume 1 | Article 242 | 3

Frontiers in Psychology | Cognition January 2011 | Volume 1 | Article 242 | 4

www.frontiersin.org January 2011 | Volume 1 | Article 242 | 5

Frontiers in Psychology | Cognition January 2011 | Volume 1 | Article 242 | 6

www.frontiersin.org January 2011 | Volume 1 | Article 242 | 7

Frontiers in Psychology | Cognition January 2011 | Volume 1 | Article 242 | 8

www.frontiersin.org January 2011 | Volume 1 | Article 242 | 9

Frontiers in Psychology | Cognition January 2011 | Volume 1 | Article 242 | 10

www.frontiersin.org January 2011 | Volume 1 | Article 242 | 11

Frontiers in Psychology | Cognition January 2011 | Volume 1 | Article 242 | 12

www.frontiersin.org January 2011 | Volume 1 | Article 242 | 13

View publication stats

You might also like