Cognition 146 (2016) 81–89

Contents lists available at ScienceDirect

Cognition
journal homepage: www.elsevier.com/locate/COGNIT

Discussion

Towards a common framework of grounded action cognition: Relating

motor control, perception and cognition
Antje Gentsch a,b, Arne Weber c, Matthis Synofzik d,e, Gottfried Vosgerau c, Simone Schütz-Bosbach b,f,⇑
a
University College London, CEHP Research Department, London, UK
b
Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany
c
Department of Philosophy, Heinrich-Heine-University Düsseldorf, Germany
d
Department of Neurodegenerative Diseases, Hertie-Institute for Clinical Brain Research, University of Tübingen, Germany
e
German Research Center for Neurodegenerative Diseases (DZNE), Tübingen, Germany
f
Psychology Department, Ludwig-Maximilians-University, Munich, Germany

a r t i c l e i n f o a b s t r a c t

Article history: The relation between motor control and action cognition – including action-related thoughts and action-
Received 8 May 2015 related perception – has been subject to controversial discussions in the last three decades. During these
Revised 11 September 2015 decades, cognitive neuroscience has been increasingly confronted with a huge variety of different
Accepted 14 September 2015
accounts trying to understand and explain the relation between these systems, their interdependencies
Available online 24 September 2015
and the mediating mechanisms by establishing notions such as ‘‘internal models”, ‘‘simulation” or
‘‘shared representation”. These accounts, however, include a large array of partly overlapping, partly con-
Keywords:
tradictory theories using similar terms for different mechanisms and different terms for similar mecha-
Grounded cognition
Motor control
nisms. In the absence of a systematic work-up and comparison, this array of accounts and theories leads
Action perception to confusion in the field, duplication of experimental work, and unconnected parallelism of theory forma-
Action cognition tion within and between different disciplines. Here we provide a systematic comparison of current mod-
Simulation els and prospective theories that deal with the relation between cognition, perception and motor control
Internal models mechanisms. In a second step, we propose ‘‘grounded action cognition” as a comprehensive metatheoret-
Common coding ical framework which defines different hypothetical possibilities of the relations between these domains,
offers systematic insights into current models and theories and last but not least may help to increase
comparability of empirical research in the domain of action and action cognition.
Ó 2015 Elsevier B.V. All rights reserved.

1. Introduction
There is a long tradition and interest in many different disci-
plines on the interplay between action and perception. This inter-
est is led by various aims: while some disciplines investigate this
interplay with the aim of understanding human ‘cognitive mecha-
nisms’ (cognitive psychology, philosophy), brain mechanisms (cog-
nitive neuroscience) or neurologic and psychiatric diseases
(neurology/psychiatry), others study this relation to model skillful
movements (robotics) or perceptual abilities (computational and
mathematical neuroscience). Some of these traditions date back
to 19th century experimental psychology and philosophical theory
of voluntary action (James, 1890/1981). Others were developed
more recently. For example, approaches from computational and
cognitive neuroscience, starting in the early 1990s, aim at describ-
⇑ Corresponding author at: Psychology Department, General and Experimental
Psychology, Ludwig-Maximilians-University Munich, 80802 Munich, Germany.
E-mail address: simone.bosbach@gmail.com (S. Schütz-Bosbach).
ing the precise brain systems and neuronal dynamics underlying
the action–perception inter-linkage by means of internal forward/-
generative models (Friston, Daunizeau, Kilner, & Kiebel, 2010;
Wolpert & Miall, 1996). The focus on the linkage between action
and perception has now been expanded to include also conceptual
abilities and cognition in general. Specifically, growing interdisci-
plinary work has now begun to relate different theoretical
approaches and empirical findings to explain also higher-
cognitive skills like mind reading in social contexts (Gallese &
Goldman, 1998) or mental imagery (Grush, 2004).
Here we take an interdisciplinary point of departure with the
aim to provide a systematic comparison of current and established
theoretical models and prospective theories that deal with the
relation between cognition, perception and motor control mecha-
nisms. We will in particular focus on the proposed internal repre-
sentational mechanisms governing mutual relations between
perception and action. Accordingly, the selection criterion for the
theories to be compared in the present paper is that they make

http://dx.doi.org/10.1016/j.cognition.2015.09.010
0010-0277/Ó 2015 Elsevier B.V. All rights reserved.
82 A. Gentsch et al. / Cognition 146 (2016) 81–89
some substantial claim about the systematic connection between
the domains of cognition, perception and motor control.1 It is
beyond the scope of the present endeavor to also take dynamic inter-
actions between mind/brain, body and world into account as pro-
posed by some more radical conceptions of embodied cognition
(for an overview compare Shapiro, 2011; or Wilson, 2002; for certain
conceptions of these radical or dynamic views see Beer, 1995; Hutto
& Myin, 2013; Keijzer, 2002). Thus, we will here specifically consider
internal representations that appear to draw on motor-related pro-
cesses. We will use the term ‘grounding’ to refer to the general rela-
tion between motor processes and action perception/cognition,
which seems underdetermined so far in current theories. Please also
note that it is not our aim to provide a systematic review of those
single theories in terms of supporting or challenging empirical evi-
dence (for a recent review see, for example, Engel, Maye, Kurthen,
& König, 2013). Instead, we here aim to systematically compare
the empirical and explanatory foci those theories adopt with respect
to the nature of the relation between motor control and action–
cognition.
This paper comprises three sections. The first provides a brief
overview of different action theories and suggests a classification
by target mechanisms. In the second section, we introduce a
framework that will be used to illustrate our notion of grounded
action cognition as a metatheoretical view. Third and finally, our
purpose is to show how existing theories can or cannot be classi-
fied into genuine grounding theories. This new classification
scheme shall offer new perspectives into commonalities and differ-
ences as well as the explanatory scope with respect to the degree
to which perceptual and cognitive abilities genuinely draw on
motor capacities; however, it is not meant to suggest an evaluation
of existing and established theories concerning their theoretical
and empirical adequacy.
2. The three main families of action theories
The large amount of highly heterogeneous, partly overlapping,
partly differing theories coming from very different disciplines is
often confusing. To facilitate a better overview and understanding,
we here classify the existing theories into three major theoretical
frameworks/families of grounded action cognition accounts: (1)
Common Coding, (2) Internal Models, and (3) Simulation theories
(Table 1; for a detailed description and analysis of each of these
families, see next paragraph). These theories operate at different
levels and interfaces between domains of action cognition and
use diverse conceptual tools (see Table 1). One shared assumption
of all major theories is that some kind of common representational
ground for action and cognition exists. This assumption has been
developed, however, at different levels of action cognition, and
with crucial differences in the empirical and explanatory focus
on motor control, perception or conceptual abilities (see Fig. 1).
This includes fundamental differences in the abilities these theo-
ries try to account for: while some theories resort to a common
ground between action and cognition to explain action planning
or motor imagery, others refer to this notion to explain the under-
standing of other individuals’ actions. The main assumptions
regarding the central thesis of grounded action cognition are sum-
marized in Table 1.
In order to understand the differences between the theories in
their attempt to link the domains of action, perception and cogni-
tion, it is helpful to consider the primary domain a theory was orig-
inally developed for and how it has been extended. For example,
1 4
For this reason, the sensorimotor theory of visual consciousness by O’Regan and
Noë (2001), for example, is not considered in the present framework: Although it
claims that visual perception is a form of action, it does not propose a specific linking
mechanism between perception and action (unlike ideomotor theory, for example).
common coding theories as well as some internal model accounts,
such as motor control theory, were explicitly developed from a
motor control point of view. Therefore, the theoretical concepts
are grounded in notions of efferent and (re-)afferent information
(i.e., motor output and sensory feedback) which have been applied
not only to understand the mechanisms behind movement kine-
matics (motor control theory) but also to describe the semantics
of action (ideomotor theory). In contrast, another version of inter-
nal models, predictive coding, was originally developed to explain
visual perception and its underlying computational architecture. It
has been extended to other sensory modalities as well as to action
in terms of active inference (the latter being a more recent devel-
opment). Hence, the key concepts in this model are rooted in
assumptions around the organization of sensory input and the per-
ception of the outside world. The ontogeny of simulation theories,
the third main family of action theories, is most heterogeneous,
since the notion of ‘simulation’ has been used to account for a vari-
ety of phenomena in the domains of action cognition and percep-
tion. For example, the notion of simulation has been used to
explain a perceptual understanding of other’s actions (mirror neu-
ron theory) and the ascription of mental states to others (simula-
tion theory), as well as the imagination of action (motor
imagery) and even abstract conceptual abilities (perceptual sym-
bols theory). Hence, understanding the primary phenomenon that
each theory is trying to account for is crucial for appreciating how
existing theories differ in relating the domains of cognition, per-
ception and motor control mechanisms. The classification of the
three families of theories according to their primary domain of
explanation is illustrated by Fig. 1.
3. Grounded action cognition: a metatheoretical view
In our approach towards a systematic framework, we will now
specify common denominators of the manifold existing assump-
tions regarding the relation between the three domains of action
cognition, perception, control and concepts of action. Current the-
ories and related experimental reports sometimes appear to use
the same terms to explain different phenomena or mechanisms,
and different terms to explain the same phenomena or similar
mechanisms. For example, the term ‘simulation’ has been used to
refer to sensorimotor processes as a reactivation of neural motor
circuitries not only when observing actions but also in the context
of reasoning about mental states of others. Likewise, the finding of
an activation of cortical areas involved in motor control during
action observation has been interpreted and labeled interchange-
ably as ‘mirroring’, ‘simulating’ or ‘mentalizing’. Moreover, terms
such as ’embodiment’, for instance, have been applied in a narrow
sense with reference to self-awareness as a form of bodily aware-
ness (i.e., the embodied self) but also in a broader sense to explain
a fully functional system of abstract concepts (cf. below, perceptual
symbol systems theory).
Moreover, study reports sometimes use vague or even over-
stated wording when reporting their findings and interpreting the
relation between the three domains. For instance, authors sympa-
thetic with simulation theories often use various and quite strong
expressions like cognition/perception (x) is ‘‘based” on a sensorimo-
tor process/the sensorimotor system (y), x ‘‘arises” in y,2 x is ‘‘local-
ized” in y,3 x is ‘‘derived” from y,4 x is ‘‘specified” by y,5 x is
2
‘‘[P]erceptual states arise in sensory-motor systems.” (Barsalou, 1999, p. 577).
3
‘‘[I]implicit memory has been localized in sensorimotor areas of the brain.”
(Barsalou, 1999, p. 589).
‘‘[A]ffordances are derived from the perceptual symbols [i.e. a systematic
activation pattern in the sensorimotor system].” (Glenberg & Kaschak, 2002, p. 599).
5
‘‘[C]onceptual processing of categories specified by motor and somatosensory
features [. . .].” (Barsalou, 1999, p. 585).
 A. Gentsch et al. / Cognition 146 (2016) 81–89 83

Table 1
Systematic overview of theories of action cognition. These theories assume a tight, but differing interrelation between the domains of motor control, action perception and action
cognition. They can be grouped into three different families of grounded action cognition accounts: Common Coding, Internal Models, and Simulation theories.

Theory Explanatory target How do motor cognition, motor control and Origin Key references
action perception relate?
Common Ideomotor Action initiation A bidirectional sensory-motor link: perceptual Cognitive Greenwald (1970) and James (1890/
Coding Theory Cognitive processes behind representations automatically activate Psychology 1981)
voluntary intentional action response representations of associated motor (motor control)
actions, and vice versa
Theory of Action planning and A common code: the representational format Cognitive Hommel et al. (2001) and Prinz (1997)
Event perception for distal perceptual contents and motor Psychology
Coding Communication between plans is identical and formed through (action
motor control and action principles of temporal coupling and perception)
perception synchronized activation
Internal Motor Motor control & motor An internal forward model which represents Computational Wolpert and Miall (1996)
Models Control learning the dynamics of the motor system, including Modeling,
Theory Implementation of optimal all sensory states resulting from motor Cognitive
online control of actions, in order to predict future states Robotics (motor
movements and posture control)
Predictive Sensory perception & motor Active inference under the free-energy Computational Friston (2005) and Friston, Mattout, and
Coding control principle: sensory surprise automatically Modeling Kilner (2011)
Perception as prediction translates into motor control representations (perception in
(similar to motor reflexes) that serve to general)
minimize sensory prediction errors by
guiding sensory sampling
Emulation Motor control & motor Draws on and combines elements of internal Philosophy Grush (2004) and Hurley (2008)
Theory imagery models and mirror mechanisms central to (action
Interrelation of neural and motor control theory and mirror neuron cognition)
representational theory
mechanisms underlying
motor control and motor
imagery
Simulation Simulation Mind reading The concept of mental simulation: action Philosophy, Gallese and Goldman (1998), Goldman
Theories Theory Ascribing mental states of observation (i.e., the perceptual Cognitive (1992), and Gordon (1986)
others by observing their representation of a motor act of others) Neuroscience
actions involves the activation of mental state (social
sequences that are isomorphic to those cognition)
underlying self-generated actions
Mirror Action observation The notion of mirror neurons: a population of Neuro- Gallese et al. (1996) and Rizzolatti,
Neurons Neuroanatomy of observed neurons coding for both observed and physiology Fadiga, Gallese, and Fogassi (1996)
actions of others executed actions, that is, the same neural (action
system is used for representing motor acts perception)
and motor percepts
Motor Mental imagery The notion of covert actions: Motor images (i. Cognitive Decety, Jeannerod, and Prablanc (1989),
Imagery Neural mechanisms e., motor percepts) are covert actions, relying Neuroscience Jeannerod (2001), Jeannerod and
underlying overt and covert on simulation of neural mechanisms and (action Decety (1995), for review’, Vogt, Rienzo,
(imagined) actions dynamics of real, physical action in low level perception) Collet, Collins, and Guillot (2013)
motor centers
Perceptual Conceptual thinking The construct of simulators: Concepts act as Cognitive Barsalou (1999) and Barsalou, Kyle
Symbols Conceptual abilities simulators by relying on a re-activation of the Psychology Simmons, Barbey, and Wilson (2003)
Theory sensorimotor states involved in experiencing (cognition in
the actual entity general)

‘‘embodied”/”grounded” in y (passim), x is ‘‘constituted” by y,6 x ‘‘re-
sides” in y,7 x/y ‘‘integrate(s)” y/x,8 y ‘‘enters” into x, y is ‘‘engaged”
in x, or y ‘‘essentially is” for x.9 All of these formulations are used to
interpret data from experimental work usually showing a correlation
between two factors. However, the fact that two factors correlate can
be interpreted in very different ways, from mere unspecified
modulation of one factor by the other to causal dependence of the
one factor on the other. Here we propose grounded action cognition
as a metatheoretical framework that aims to define hypothetical
possibilities for the relations between different domains of action
representation in order to provide a systematic synopsis and a
6
‘‘[O]ur ability to perceive not only depends on, but is constituted by, our
possession of [. . .] sensorimotor knowledge.” (Noë, 2004, p. 2).
7
‘‘[C]ognitive and perceptual representations reside in a common system.”
(Barsalou, 1999, p. 589).
8
‘‘[T]he sensorimotor integration supported by the premotor-parietal F5-PF
mirror-matching system [. . .] also to provide a meaningful account of the goals and
purposes of others’ actions, by means of their simulation.” (Gallese, 2003, p. 523).
9
‘‘[M]otor behavior of an exploring animal is essential to perception [. . .].” (Hurley,
2001, p. 11).
path towards an integrative understanding of current theories of
action. Within this new framework, a crucial distinction is drawn
between constitution relation and acquisition relation, which offers
a systematic classification tool for existing and future theories of
grounded action cognition.
3.1. Acquisition and constitution: specifying the relation between
perception/cognition and action
In order to specify the claim of grounded cognition, it is crucial
to scrutinize the exact relation that is assumed to exist between
two domains of interest (e.g., between motor control and action
perception, or action cognition). So far, the question of whether
grounded or embodied cognition is meant to oppose ‘‘classical”
modular pictures of the mind (Fodor, 1975; Pylyshyn, 1984;
Smith & Medin, 1981; Tulving, 1972) is in fact not answered. There
is certainly no doubt that action (i.e., motor control) and percep-
tion/cognition are somehow interrelated. The question is thus
not whether these processes are related, but how they are related.
Representatives of the ‘‘classical” picture of the mind even do not
84 A. Gentsch et al. / Cognition 146 (2016) 81–89

Fig. 1. Theories of ‘‘grounded action cognition” and their main domain of focus. The diagram relates theoretical accounts of action cognition and classifies them according to
their primary empirical and explanatory focus. All theories draw on a relation between action and cognition to explain a mental ability related to motor action. However, they
differ in the domain and process of action cognition they primarily refer to (action control, perception or concepts), and in the mental ability they are trying to account for.
Simulation Theories (green) are concerned with action perception and conceptual representations in order to understand socio-cognitive skills of mind-reading based on
mechanisms of motor control. Internal Models (blue) aim at describing primarily the interrelation between motor control and perception of action, by assuming that
perception arises with reference to motor control functions (cf. notions of ‘‘forward models” and ‘‘prediction error”) and motor planning arising with reference to perception
(cf. notions of ‘‘inverse models” or ‘‘active inference”). Common Coding (orange) accounts span all three areas of action perception, control and concepts by explaining (late)
event perception and (early) action planning through the existence of a shared sensorimotor representational code at a higher level.

explicitly deny a perception–action interrelation (although they
usually do not focus on it), but they typically claim this relation
to be rather weak, for example in the sense of mere modulating
bi-directional influences. Thus, our reconstruction of the percep-
tion–action interrelation meant by the different authors above will
be such that their claims are indeed opposing more classical views
in claiming a much stronger, thus functional determining relation
between action and perception in the sense of acquisition and/or
constitution. In a first step, we will now introduce the technical
terms of acquisition conditions and of constitution, respectively
(cf., Weber & Vosgerau, 2012). We will illustrate how we under-
stand these terms with the help of empirical examples and we will
also discuss some empirical implications that follow from this new
perspective.
Acquisition can be illustrated by a classical experiment of Held
and Hein (1963) in newborn animals. They divided kittens into two
groups; kittens in one group pull a carriage around a room in a hor-
izontally rotating carousel, while kittens in the other group ride
this carriage. The carousel was set up such that both groups of kit-
tens have the same visual stimuli, the only difference being that
kittens in the first group are allowed to make active movements
while the kittens in the other group were only passively moved.
After spending some time in the carousel, tests show that the rid-
ing kittens did not develop normal perceptual abilities as for exam-
ple depth perception. Obviously, these kittens were not able to
integrate their visual experiences with movements of their own
bodies, which caused visual deficits in the long run (see also,
Gibson, 1969). These findings suggest that the ability to actively
move (ability A) is among the acquisition conditions for the ability
to perceive depth (ability B). However, once B is acquired on the
basis of A, ability A can get lost without any negative impact on
ability B. So, a kitten that already acquired normal sight will not
lose its ability to perceive if we put it in a carriage (thus hindering
it from active movement). Active movement is only necessary to
acquire the ability to perceive but not necessary to maintain this
capacity once it is fully acquired.
‘‘Constitution”, in contrast, refers to a much tighter interpreta-
tion of the relation between action and perception. An almost triv-
ial example to illustrate a real constitution relation is that the
ability to move one’s own legs is constitutive for the ability to walk,
while both abilities are not identical, since it is possible to be able
to move one’s own legs without being able to walk. So, already in
this example it is obvious that constitution does not necessarily
mean that there is no more to ability B than just ability A: indeed,
there is much more to the ability to walk than just the ability to
move one’s own legs. Nevertheless, you cannot do the latter with-
out being able to do the former. However, constitutive relations
can also be interpreted less strong. For example, if the complete
loss of A only leads to some impairments of B, it is still valid to
argue that A is a constitutive element of full-blown ability B. Thus,
although the criterion for constitutive relations might read very
strong at first sight, it only requires a systematic dependence
between break-downs in either ability.
So far only relatively few10 empirical studies were able to indi-
cate that motor control processes indeed play a constitutive role in
10
Among notable exceptions are, for example, studies using brain stimulation such
as Transcranial Magnetic Stimulation (TMS; Moore, Ruge, Wenke, Rothwell, &
Haggard, 2010, linking pre-SMA to both action preparation and action perception)
or invasive, intracranial techniques such as direct electrical stimulation (DES;
Desmurget et al., 2009, linking parietal cortex to both representing motor intentions
and forward modeling for movement monitoring). Although these studies attempt to
show causal relations, however, even these methods have some caveats. The effects
evoked by surface stimulation of the cortex are complex, mostly involving a large
volume of brain tissue, and behavioral effects across studies are heterogeneous
ranging from facilitation to inhibition of responses. For example, TMS stimulation of a
specific target area may induce effects in a larger network of interconnected brain
regions depending on different physiological and morphological factors. Thus, these
techniques may not allow us to draw unequivocal conclusions about whether an
effect is indeed being generated in a stimulated brain area.
 A. Gentsch et al. / Cognition 146 (2016) 81–89
action cognition, thus going beyond mere correlations between
motor control processes and action cognition. A prominent example
is a study that investigated patients diagnosed with amyotrophic lat-
eral sclerosis (ALS). ALS refers to a neurodegenerative condition that
is defined by progressive degeneration of motor tracts. This study
showed that ALS patients were significantly more impaired in con-
cepts of actions than in concepts of objects, and that this difficulty
specifically correlated with atrophy of the motor cortex (Grossman
et al., 2008). These findings suggest that intact conceptual knowl-
edge of action features depends on an intact motor cortex. If this
finding can be corroborated by future studies, confirming indeed a
specific relation between degraded action knowledge and degraded
motor systems (and not just degraded action knowledge as part of
the more widespread extramotor cortical atrophy known in ALS
(Libon et al., 2012), this would suggest that motor systems are
indeed constitutive for action cognition. Intact motor systems would
here present the structural presupposition to intact action cognition.
In another domain, intact motor processes present the func-
tional presupposition for acquiring intact perception. Microsac-
cades, for example, allow restoring faded vision during fixation,
for both foveal and peripheral targets. Accordingly, it has been
shown already in the 1950s that when eye movements are com-
pletely suppressed (for instance, by techniques of retinal stabiliza-
tion (Ditchburn & Ginsborg, 1952), our perception of stationary
objects appears to fade completely (Martinez-Conde, Macknik,
Troncoso, & Dyar, 2006; for review see, Martinez-Conde, Otero-
Millan, & Macknik, 2013). For the more specific thesis that motor
processes are indeed constitutive elements for acquiring intact
action perception, it would be necessary to demonstrate that sim-
ilar process are at work in the perception of actions.
3.2. The partial constitution criterion
Our proposal here is that the only genuine grounding relation is
that of a partial constitution relation between perception and
action, in the sense that there is more to ability B than just ability
A. Thus, on this view the relation does not collapse into an identity
relation. To make this argument clearer, let us take the hypotheti-
cal case of a radical and full constitution of a certain perceptual/
cognitive ability by motor processes. This would lead us to the
assumption that this ability is just a subclass of or even identical
to some (complex) motor ability. Moreover, if all perceptual/cogni-
tive abilities would be so fully constituted by motor processes,
some basic motor abilities would be constitutive of all action-
perceptual and -cognitive capacities, and this implies that the
two domains would fall together and become the same. However,
if perceptual processes or states are just kinds of motor processes
or states, there would be no intelligible sense of understanding a
Fig. 2. Schematic representation of different types of constitution relation between two abilities, A and B. (i): Some B-ability correlates with some A-ability (dotted line),
while there is no constitution relation between the two. (ii): A B-ability is partially constituted by A-abilities if the loss of A leads to severe impairment of the B-ability but not
to a complete breakdown (since there is more to ability B than just A). (iii): The B-ability is fully constituted by A-abilities if there is no non-A-ability constitutive for it, and in
this case the B-ability is nothing but a (complex) A-ability itself.
85
certain perceptual process as ‘‘grounded” in another basic process,
since then there is nothing left to be ‘‘grounded”; the former pro-
cess would be merely identical to the latter. Moreover, the differ-
ent domains of motor control and action perception would not
be discriminable in empirical research. The other hypothetical
extreme case would be that there is no constitution at all of one
process by another. This would then in the end lead to a picture
similar to the classical conception of a modular mind (see above).
Then, however, there would be no novel and challenging way of
talking about some cognitive processes potentially being
grounded. Such relations would simply be too weak to establish
a picture opposing the classical one. Therefore, we argue that
grounded action cognition should be understood and specified in
terms of a partial constitution relation. The different types of con-
stitution relation are illustrated by Fig. 2.
4. Action cognition theories and implied constitution relation
In the first part of the present paper, we organized the estab-
lished action cognition theories into three major families, Common
Coding (CC), Internal Models (IM) and Simulation Theories (ST). We
will now analyze them in more detail with respect to whether they
can be considered genuine grounding theories; that is, they will be
compared according to the implied constitution relation between
motor control and action cognition. The aim shall be to put the
variety of existing accounts of action into a systematic frame of
grounded action cognition. Each family contains some theories that
implicitly assume an at least partial constitutive relation between
action and perception, while other theories in contrast do not
imply a constitutive relation. In Table 2, we made an attempt to
organize those theories according to the constitution relation they
assume to hold. We will discuss the categorization of each theory
family in our table in turn.
4.1. Common Coding Theory
Common Coding theories are characterized by the assumption
that there is a bidirectional or common sensorimotor code. The
ideomotor theory assumes that action and perceptual representa-
tions are separate representations in principle, but can be co-
activated to generate and facilitate voluntary action and adaptive
control of goal-directed behavior. Thus, the motor domain is here
not seen as constitutional for perceptual and cognitive abilities
(see Table 2). In contrast, the theory of event coding holds that
there is a common representational code for action and perception.
This code structures action planning and action perception
and therefore action and perception are related through this
86 A. Gentsch et al. / Cognition 146 (2016) 81–89
Table 2
The three major families of action cognition theories and the main assumptions regarding conditions of constitution.
Constitution Theory
relation
Common Coding No Ideomotor theory
No Theory of event coding
Internal Models Partial Motor control theory
Partial Predictive coding
No Emulation theory
Simulation Full Mirror neuron theory;
Theories Motor imagery
Full Perceptual symbols theory
representational format. Critically, however, the code is indepen-
dent of executive action domains as well as of sensory modalities.
Thus, following this picture, the domain of motor control is not
directly connected to (or even constitutional for) sensory modali-
ties. Instead, a common code enables an indirect communication
between the two domains. Moreover, motor control is not seen
as the primary system because the perceptual system and the
motor control system are coequal.
The main aim of Common Coding theories is to explain the rep-
resentations of action goal and action effects (a common represen-
tational domain for perception and action) with the cognitive
antecedent of the action (prospective coding in event representa-
tion’; Schütz-Bosbach & Prinz, 2007). The main thesis is that fea-
tures relevant for motor control are involved already at the level
of action perception and the perception of action effects, with
abundant empirical evidence, for example from classical studies
on spatial stimulus–response compatibility effects (e.g., Elsner &
Hommel, 2001; Hommel, Musseler, Aschersleben, & Prinz, 2001;
Kunde, 2001; see Shin, Proctor, & Capaldi, 2010 for a review). The
typical observation is an advantage (in terms of speed and accu-
racy) of responses that share ‘‘motoric” features with a target stim-
ulus (such as the relative location in space), supposedly due to the
automatic activation of a common location code between stimulus
and response. An elaborated version of the theory of event coding
assumes a common format of (late) perception and (early) motor
control which are coded in a common representational medium,
i.e. event codes that are both motoric and perceptual, commonly
coded, ‘‘especially if the features of perceived and to-be-
produced events overlap. [. . .] In other words, perceptual and
action-planning processes only interact if the codes they operate
on refer to the same (kind of) feature of a distal event” (p. 862’,
Hommel et al., 2001).
In contrast to Simulation Theories (as in the case of mirror neu-
rons), Common Coding – at least in its original version – is not a
framework for understanding actions of others but mainly for
action planning. In contrast to Internal Model theories, action con-
trol in Common Coding does not draw on a forward signal which is
sent in parallel with the motor command to guarantee optimal
action performance. Whereas Common Coding emphasizes percep-
tion because it supposedly shapes motor control (at least with
respect to distal events) Internal Model theories focus primarily
on motor control. To summarize, Common Coding assumes three
different classes of representational formats: purely motoric,
purely sensory and a common code format. Thus, neither motor
control is grounded in perception nor the other way round, but
both are linked through additional (common code) representa-
tions. None of the Common Coding theories can therefore be con-
sidered, from a constitution point of view, as a genuine grounding
theory.
Central hypothesis
Separate motor and perceptual representations can be co-activated
A common code is assumed, but only for representation of distal events and in addition to lower-
level sensory and motor representations
Internal forward models are thought to run in parallel to motor control processes
In probabilistic models, processes of motor control and perceptual inference are identical upwards
of the spinal level
Cognition (including motor imagery and motor planning) is based on the internal emulation of
motor and sensory processes
Specific, separate subpopulation of neurons in motor areas with mirror or simulation properties
Simulation requires a reenactment of sensorimotor mechanisms, no additional process is assumed
4.2. Internal Models
Internal Model theories are characterized by the proposed exis-
tence of internal models of the motor apparatus. These models are
postulated neural processes that extrapolate the commands of the
motor system in order to estimate and anticipate the outcome of a
motor command. One variant of Internal Model theories, motor
control theory, has its origin in control theory and robotics and pri-
marily aims at explaining optimal online control of movements
such as posture control and movement trajectory planning. More
precisely, motor control theory assumes different internal models:
a forward internal model predicts the sensory consequences from
efference copies of issued motor commands; additionally, an
inverse model calculates the necessary feedforward motor com-
mands from the desired movement trajectory (cf. Kawato, 1999).
To explain perception, especially of self-generated actions, forward
models are sometimes also simply referred to as ‘‘efference copy”
(von Holst & Mittelstaedt, 1950) or ‘‘corollary discharge” (Bell &
Grant, 1989; Sperry, 1950). Following this interpretation, percep-
tion and action are essentially dependent on motor intentions
and related motor signals, such as the efference copy and predic-
tions computed on their basis. Extensive evidence for this depen-
dency comes from studies on the phenomenon that people
cannot tickle themselves, for which predictions have been attribu-
ted a prominent role by attenuating the sensory consequences of
self-generated actions (Blakemore, Frith, & Wolpert, 1999). How-
ever, the internal model itself, despite being a decisive part of
motor control, is not identical with the motor mechanism, other-
wise a comparison between a predicted and actual state for opti-
mizing motor control (e.g. fast online corrections of movement
trajectories) would not be possible. That is, motor control theory
postulates that there is one input signal (i.e. desired action out-
come) and two different output processes (i.e. motor command
and efference copy), and motor control is emphasized as a consti-
tutive condition. Therefore, according to our suggested framework,
a constitution of perception by motor control mechanisms can be
found but this constitution is not exhaustive (i.e. there are also
other factors involved). Hence, in motor control theory, some
motor representations are considered partially constitutive at least
for parts of action cognition.
Another type of Internal Model theory, predictive coding, is
based on the view of perception as prediction (Friston et al.,
2010). In contrast to motor control theory, the primary focus of this
approach is on perceptual models and not on motor control mech-
anisms. However, active inference as a corollary of this perspective
also aims to account for many aspects of motor behavior, from ocu-
lomotor reflexes to goal-directed movements. It assumes a gener-
ative forward model which maps from motor control to sensory
consequences, and which forms part of a larger hierarchical model
 A. Gentsch et al. / Cognition 146 (2016) 81–89
used for perceptual inference in general. In contrast to motor con-
trol theory, it is thought to serve all aspects of motor control with-
out the need for a separate inverse model to provide control signals
(cf Friston, 2005). In this respect, forward-inverse models in motor
control are formally distinct from predictive coding accounts of
action perception. In predictive coding accounts, motor commands
are not constructed by an inverse model but arise from an inver-
sion of the forward model. That is, instead of a mapping from con-
trol signals to sensory consequences and vice versa, motor control
signals are identical with descending sensory predictions of propri-
oceptive sensations. However, beyond internal proprioceptive con-
sequences of movements, exteroceptive consequences of actions
(e.g., vision, audition) are considered corollary discharges, as in
classical forward models. Thus, at least for exteroceptive modali-
ties of action perception in predictive coding mechanisms, the
same holds true as for motor control theories: some motor repre-
sentations are constitutive for parts of action cognition.
Some types of Internal Model theories incorporate elements
from Simulation theories. For example, the emulation theory, as
proposed e.g. by Rick Grush (2004), combines constructs like effer-
ence copy and motor prediction with mirror mechanisms. It seeks
to explain how motor control operations in classical motor centers
can produce motor imagery and certain aspects of visual percep-
tion by driving the operation of a motor-visual emulator of the
musculoskeletal system. Beyond the motor domain, the theory
holds that imagery in general is based on both amodal and
modality-specific systems driving parallel emulators as real time
models of behavior and perception. Hence, the theory has been
applied not only as an explanatory model of motor control and
motor imagery, but also of several other cognitive functions, indi-
cating dissimilarity between motor processes simpliciter and cogni-
tive functions. Accordingly, in Grush’s discussion of emulation
versus simulation accounts of motor imagery, he insists that ‘‘mere
operations of the motor center is not enough; to produce imagery
the motor centers must be driving an emulator of the body” and ‘‘
[a] motor plan is one thing, a sequence of proprioception and
kinesthesis is another” (p. 385’, Grush, 2004). Thus, according to
Grush, the executed action and its effects are not sufficient (and
maybe not even necessary) to install action-related perception
and cognition.
Taken together, all types of Internal Model theories postulate
that action perception and cognition are significantly informed
and construed by predictions. In particular, exteroceptive predic-
tions are not identical with motor commands but are grounded
in the latter. However, more than motor commands is necessary
to produce these predictions, namely sensory feedback, context
cues, and context estimates. Thus, predictions and related action
perception/cognition cannot be exhaustively grounded in motor
commands. Therefore, according to Internal Model theories, motor
abilities would be (at most) partially constitutive of action-related
perceptual and cognitive capacities.
4.3. Simulation Theories
The family of Simulation Theories is characterized by the gen-
eral assumption that cognition is essentially simulation-like (re-)
use of the brain’s modal systems, such as the sensorimotor system.
For example, the mirror neuron theory assumes a population of
neurons coding for both observed and executed actions, that is,
the same neural sub-system is used for representing motor acts
and motor percepts (for review see, Rizzolatti & Craighero, 2004).
This applies specifically to so-called ‘strictly congruent’ mirror
neurons (Gallese, Fadiga, Fogassi, & Rizzolatti, 1996), encoding
the kinematics of an action, in contrast to some class of ‘broadly
congruent’ mirror neurons (or ’logically-related’ mirror neurons;
e.g. Iacoboni et al., 2005) which encode action goals with a less
87
strict correspondence between motor control and action percep-
tion. It should be noted that mirror neurons are regarded as neuro-
physiological instantiations of simulation and, therefore, mirror
neuron theories provide a mechanistic explanation of the involved
neuronal processes, rather than offering a theoretical account on
their own.
Similarly, the motor imagery account states that motor per-
cepts (or ‘‘motor images”) are ‘‘covert actions”, i.e. a simulation
of neural mechanisms and dynamics of real, physical action in
low level motor centers before an action is executed (Jeannerod,
2001). Here, Jeannerod (2001, p. 106) argues that ‘‘lesions affecting
structures located ahead of motor system [. . .] affect the ability to
produce a given S-state”. By the term S-state, Jeannerod refers to
‘‘those ‘mental states’ which involve an action content and where
brain activity can be shown to simulate that observed during the
same, executed action” (Jeannerod, 2001, p. 103). But these simu-
lations are ‘‘covert actions” that do not result in muscular activity
as an overt movement.
Another prominent example of Simulation Theories is the the-
ory of perceptual symbol systems (Barsalou, 1999). According to
this theory, cognition exclusively consists of activation patterns
of various sensory modalities, here the perceptual symbol system.
In this case cognitive and perceptual mechanisms share the same
representational states because cognitive processing is essentially
a simulation, i.e. a re-activation of sensory states, including senso-
rimotor activation patterns underlying the original perceptual
state. This reactivation of the same neural circuits involved during
perception of and action towards those objects allows the re-
enactment of multimodal information. Consequently, also higher-
order abilities such as thinking or categorization are ‘‘grounded”
in low-level sensorimotor abilities. Empirical evidence for the the-
ory of perceptual symbol systems indicates that perceptual and
motor mechanisms are engaged when a cognitive system performs
conceptual processing. It is suggested that, for example, talking or
thinking about objects actually (partly) consists of the reactivation
of previous experiences stored in the perceptual symbol system
(for an overview cf. Barsalou, 2008; de Vega, Glenberg, &
Graesser, 2008; Pecher & Zwaan, 2005; Semin & Smith, 2008).
Based on mirror neuron theories, even cognitive abilities such
as mind-reading and understanding mental states of other’s by
observing their actions are often ‘‘explained” by simulation mech-
anisms that shall help ‘‘put oneself in the ‘mental shoes’ of others”
(cf. Gallese & Goldman, 1998).
All variants of simulation theories understand sensorimotor
processes as being primary (in contrast to Common Coding). More-
over, in contrast to Internal Models, they focus on higher order cog-
nitive abilities to explain conceptual thinking and mind-reading
underlying social understanding and social communication, rather
than focusing solely on motor control functions. Following Simula-
tion Theories in general, cognitive abilities such as motor imagery,
conceptual abilities or understanding the mental states of others
rely on a reactivation of sensorimotor states. Because simulation
is essentially a reenactment of sensorimotor modalities there is
nothing over and above those sensorimotor mechanisms. Thus,
simulation theories postulate an identity relation (i.e., full consti-
tution) between motor abilities and action cognition. A meaningful
use of the notion of ‘grounded cognition’, however, can only be
based on a relation of constitution and not of an identity between
the relevant domains of action cognition, perception and motor
control, as claimed by simulation theories.
5. Conclusions
Following the modular picture of the mind, cognition and action
has commonly been assumed to operate in strictly different
88 A. Gentsch et al. / Cognition 146 (2016) 81–89
domains: cognition draws on abstract conceptual representations,
whereas motor control functions are thought to rely on relatively
low-level and automatic processes. In recent years, however, cog-
nitive neuroscience has been confronted with an increasingly large
array of partly overlapping, partly contradicting theories that
diverge broadly with respect to the degree to which cognition is
thought to be grounded in systems for perception and action. Con-
sequently, the proposed conceptual relations of motor control and
action cognition are manifold. This has led to apparent duplication
of experimental work, a relative huge collection of unsorted exper-
imental findings and unconnected development of different theo-
ries. Recently, there have been some attempts comparing and
linking different theories pairwise. For example, connections have
been made between theories belonging to the family of Internal
Models and Simulation Theory (Kilner, Friston, & Frith, 2007),
between Common Coding accounts and Simulation Theory
(Iacoboni, 2009), or between Common Coding and Internal Models
(Chambon & Haggard, 2013). However, up to date, a comprehen-
sive overall framework able to accommodate all the theories is still
missing. We here argued that ‘‘grounded action cognition” investi-
gating constitution conditions between motor processes and action
cognition may provide a comprehensive metatheoretical frame-
work that allows bringing together this heterogeneous mix of the-
ories and to analyze and classify them according to the specific
relation they claim to hold between motor processes and action
perception/action cognition. Thereby, we do not evaluate the dif-
ferent views according to their theoretical and empirical adequacy
as such, but we do evaluate them according to their claim that
there is a grounding relation between motor processes and action
perception/cognition that is denied by ‘‘classical” theories. Thus,
some of the theories turn out to be genuine grounding theories,
while others do not qualify as grounding theories. Theories that
do not qualify as grounding theories despite of their wording do
so either because they claim a grounding relation that is too weak
to contradict classical theories, or because they essentially claim
that the domains are identical, and thus overstretch the relation
for grounding.
Of the nine different action theories, which we classified into
three different families based on their primary explanatory scope,
only very few can be considered genuine grounding theories.
According to Common Coding accounts, separate representational
codes are assumed which mediate between action and perception
and thereby subserve both the control and evaluation of actions.
However, acting and perceiving are seen as independent from each
other, hence they are without any constitutive function for the
other domain. Internal Model theories take the grounding relation
one step further by assuming that perceived action events are
essentially construed by predictions. These predictions are thought
to be grounded in motor commands, but they are also informed by
other types of contextual cues and prior knowledge. Thus, implicit
in the idea of internal models is the assumption that motor abilities
are partially constitutive of action-related perceptual and cognitive
capacities. Within the family of Simulation Theory accounts, the
classification depends on the specific version. For example, the
mirror-neuron version supposes an identity relation between
motor control and perceptual processes in the early stages of pro-
cessing. Similarly to Internal Model theories, the emphasis is on
motor control mechanisms, however, these are understood as fully
constitutive for some perceptual processes. In more general ver-
sions of Simulation Theory – like Barsalou’s perceptual symbol sys-
tem – the sensorimotor system plays a constitutive role for all
cognitive and perceptual processing, the constitution being
exhaustive.
In summary, considering action theories from a grounded cog-
nition perspective provides a systematic tool for evaluating theo-
retical assumptions on how perception and action are
represented and shape one another. This may help to determine
adequate empirical approaches to core questions around embodied
human cognition. For example, to which extent can motor control
deficits or motor constraints affect the perception of other’s actions
and socio-cognitive functions at higher levels? How does motor
expertise or training determine action-related perceptual content?
Which aspects of self-awareness in the domain of action, such as
feelings and judgments of self-agency, are reducible to certain
motor control functions? The present framework of constitution
and constitution relations opens a new perspective for specifying
the relation between motor processes and action cognition and
may help to find the right balance between reductionist point of
views on the one hand and purely associative theories on the other
hand. It may help to further discriminate the different domains of
action perception and motor states as well as gain further insights
in common factors and interdependences of different domains.
Acknowledgments
This project was supported by a grant of the Volkswagen Stif-
tung to all of the authors (‘‘Grounding thoughts in actions, Thin-
kAct’’; Az. II/85 068/85 158/85 159) and a Heisenberg fellowship
(SCHU 2471/4-1) of German Research Foundation/Deutsche
Forschungsgemeinschaft (DFG) to SS-B.
References
Barsalou, L. W. (1999). Perceptual symbol systems. Behavioral and Brain Sciences, 22
(4), 577–609. Discussion 610-560.
Barsalou, L. W. (2008). Grounded cognition. Annual Review of Psychology, 59,
617–645.
Barsalou, L. W., Kyle Simmons, W., Barbey, A. K., & Wilson, C. D. (2003). Grounding
conceptual knowledge in modality-specific systems. Trends in Cognitive Sciences,
7(2), 84–91.
Beer, R. D. (1995). A dynamical systems perspective on agent-environment
interaction. Artificial Intelligence, 72(1–2), 173–215.
Bell, C. C., & Grant, K. (1989). Corollary discharge inhibition and preservation of
temporal information in a sensory nucleus of mormyrid electric fish. Journal of
Neuroscience, 9(3), 1029–1044.
Blakemore, S. J., Frith, C. D., & Wolpert, D. M. (1999). Spatio-temporal prediction
modulates the perception of self-produced stimuli. Journal of Cognitive
Neuroscience, 11(5), 551–559.
Chambon, V., & Haggard, P. (2013). Premotor or ideomotor: How does the
experience of action come about? In W. Prinz, M. Beisert, & A. Herwig (Eds.),
Action science: Foundations of an emerging discipline (pp. 359–380). MIT Press.
de Vega, M., Glenberg, A. M., & Graesser, A. C. (2008). Symbols and embodiment:
Debates on meaning and cognition. Oxford: Oxford University Press.
Decety, J., Jeannerod, M., & Prablanc, C. (1989). The timing of mentally represented
actions. Behavioural Brain Research, 34(1–2), 35–42.
Desmurget, M., Reilly, K. T., Richard, N., Szathmari, A., Mottolese, C., & Sirigu, A.
(2009). Movement intention after parietal cortex stimulation in humans.
Science, 324(5928), 811–813.
Ditchburn, R. W., & Ginsborg, B. L. (1952). Vision with a stabilized retinal image.
Nature, 170(4314), 36–37.
Elsner, B., & Hommel, B. (2001). Effect anticipation and action control. Journal of
Experimental Psychology: Human Perception and Performance, 27(1), 229–240.
Engel, A. K., Maye, A., Kurthen, M., & König, P. (2013). Where’s the action? The
pragmatic turn in cognitive science. Trends in Cognitive Sciences, 17(5), 202–209.
Fodor, J. A. (1975). The language of thought. Harvard University Press.
Friston, K. (2005). A theory of cortical responses. Philosophical Transactions of the
Royal Society of London, Series B: Biological sciences, 360(1456), 815–836.
Friston, K., Daunizeau, J., Kilner, J., & Kiebel, S. J. (2010). Action and behavior: A free-
energy formulation. Biological Cybernetics, 102(3), 227–260.
Friston, K., Mattout, J., & Kilner, J. (2011). Action understanding and active inference.
Biological Cybernetics, 104(1–2), 137–160.
Gallese, V. (2003). The manifold nature of interpersonal relations: The quest for a
common mechanism. Philosophical Transactions of the Royal Society of London,
Series B: Biological sciences, 358(1431), 517–528.
Gallese, V., Fadiga, L., Fogassi, L., & Rizzolatti, G. (1996). Action recognition in the
premotor cortex. Brain, 119(Pt 2), 593–609.
Gallese, V., & Goldman, A. (1998). Mirror neurons and the simulation theory of
mind-reading. Trends in Cognitive Sciences, 2(12), 493–501.
Gibson, E. J. (1969). Principles of perceptual learning and development. New York:
Appleton-Century Crofts.
Glenberg, A. M., & Kaschak, M. P. (2002). Grounding language in action. Psychonomic
Bulletin & Review, 9(3), 558–565.
Goldman, A. I. (1992). In defense of the simulation theory. Mind & Language, 7,
104–119.
 A. Gentsch et al. / Cognition 146 (2016) 81–89
Gordon, R. M. (1986). Folk psychology as simulation. Mind & Language, 1, 158–171.
Greenwald, A. G. (1970). Sensory feedback mechanisms in performance control:
With special reference to the ideo-motor mechanism. Psychological Review, 77
(2), 73–99.
Grossman, M., Anderson, C., Khan, A., Avants, B., Elman, L., & McCluskey, L. (2008).
Impaired action knowledge in amyotrophic lateral sclerosis. Neurology, 71(18),
1396–1401.
Grush, R. (2004). The emulation theory of representation: Motor control, imagery,
and perception. Behavioral and Brain Sciences, 27(3), 377–396. Discussion 396-
442.
Held, R., & Hein, A. (1963). Movement-produced stimulation in the development of
visually guided behavior. Journal of Comparative and Physiological Psychology, 56,
872–876.
Hommel, B., Musseler, J., Aschersleben, G., & Prinz, W. (2001). The theory of event
coding (TEC): A framework for perception and action planning. Behavioral and
Brain Sciences, 24(5), 849–878. Discussion 878-937.
Hurley, S. (2001). Perception and action: Alternative views. Synthese, 129, 3–40.
Hurley, S. (2008). The shared circuits model (SCM): How control, mirroring, and
simulation can enable imitation, deliberation, and mindreading. Behavioral and
Brain Sciences, 31(1), 1–22. Discussion 22-58.
Hutto, D. D., & Myin, E. (2013). Radicalizing enactivism: Basic minds without content.
Cambridge, Massachusetts: MIT Press.
Iacoboni, M. (2009). Imitation, empathy, and mirror neurons. Annual Review of
Psychology, 60, 653–670.
Iacoboni, M., Molnar-Szakacs, I., Gallese, V., Buccino, G., Mazziotta, J. C., & Rizzolatti,
G. (2005). Grasping the intentions of others with one’s own mirror neuron
system. PLoS Biology, 3(3), e79.
James, W. (1890/1981). The principles of psychology. Harvard University Press
(original work published 1890).
Jeannerod, M. (2001). Neural simulation of action: A unifying mechanism for motor
cognition. Neuroimage, 14(1 Pt 2), S103–S109.
Jeannerod, M., & Decety, J. (1995). Mental motor imagery: A window into the
representational stages of action. Current Opinion in Neurobiology, 5(6),
727–732.
Kawato, M. (1999). Internal models for motor control and trajectory planning.
Current Opinion in Neurobiology, 9(6), 718–727.
Keijzer, F. (2002). Representation in dynamical and embodied cognition. Cognitive
Systems Research, 2002(3), 275–288.
Kilner, J. M., Friston, K. J., & Frith, C. D. (2007). Predictive coding: An account of the
mirror neuron system. Cognitive Processing, 8(3), 159–166.
Kunde, W. (2001). Response-effect compatibility in manual choice reaction tasks.
Journal of Experimental Psychology: Human Perception and Performance, 27(2),
387–394.
Libon, D. J., McMillan, C., Avants, B., Boller, A., Morgan, B., Burkholder, L., et al.
(2012). Deficits in concept formation in amyotrophic lateral sclerosis.
Neuropsychology, 26(4), 422–429.
Martinez-Conde, S., Macknik, S. L., Troncoso, X. G., & Dyar, T. A. (2006).
Microsaccades counteract visual fading during fixation. Neuron, 49(2), 297–305.
89
Martinez-Conde, S., Otero-Millan, J., & Macknik, S. L. (2013). The impact of
microsaccades on vision: Towards a unified theory of saccadic function.
Nature Reviews Neuroscience, 14(2), 83–96.
Moore, J. W., Ruge, D., Wenke, D., Rothwell, J., & Haggard, P. (2010). Disrupting the
experience of control in the human brain: Pre-supplementary motor area
contributes to the sense of agency. Proceedings of the Royal Society of London B:
Biological Sciences, 277(1693), 2503–2509.
Noë, A. (2004). Action in perception. Cambridge, MA: MIT Press.
O’Regan, J. K., & Noë, A. (2001). A sensorimotor account of vision and visual
consciousness. Behavioral and Brain Sciences, 24(5), 939–973. Discussion 973-
1031.
Pecher, D., & Zwaan, R. A. (2005). Grounding cognition: The role of perception and
action in memory, language and thinking. Cambridge: Cambridge University
Press.
Prinz, W. (1997). Perception and action planning. European Journal of Cognitive
Psychology, 9(2), 129–154.
Pylyshyn, Z. W. (1984). Computation and cognition. MIT Press.
Rizzolatti, G., & Craighero, L. (2004). The mirror-neuron system. Annual Review of
Neuroscience, 27, 169–192.
Rizzolatti, G., Fadiga, L., Gallese, V., & Fogassi, L. (1996). Premotor cortex and the
recognition of motor actions. Brain Research. Cognitive Brain Research, 3(2),
131–141.
Schütz-Bosbach, S., & Prinz, W. (2007). Prospective coding in event representation.
Cognitive Processing, 8(2), 93–102.
Semin, G. R., & Smith, E. R. (2008). Embodied grounding: Social, cognitive, affective,
and neuroscientific approaches. New York: Cambridge University Press.
Shapiro, L. (2011). Embodied cognition. New York: Routledge Press.
Shin, Y. K., Proctor, R. W., & Capaldi, E. J. (2010). A review of contemporary
ideomotor theory. Psychological Bulletin, 136(6), 943–974.
Smith, E. E., & Medin, D. L. (1981). Categories and concepts. Harvard University Press.
Sperry, R. W. (1950). Neural basis of the spontaneous optokinetic response
produced by visual inversion. Journal of Comparative and Physiological
Psychology, 43(6), 482–489.
Tulving, E. (1972). Episodic and semantic memory. In E. Tulving & W. Donaldson
(Eds.), Organization and memory (pp. 381–403). Academic Press.
Vogt, S., Rienzo, F. D., Collet, C., Collins, A., & Guillot, A. (2013). Multiple roles of
motor imagery during action observation. Frontiers in Human Neuroscience, 7,
807.
von Holst, E., & Mittelstaedt, H. (1950). Das Reafferenzprinzip. Naturwissenschaften,
37, 464–476.
Weber, A. M., & Vosgerau, G. (2012). Grounding action representations. Review of
Philosophy and Psychology, 3(1), 53–69.
Wilson, M. (2002). Six views of embodied cognition. Psychonomic Bulletin & Review,
9(4), 625–636.
Wolpert, D. M., & Miall, R. C. (1996). Forward models for physiological motor
control. Neural Networks, 9(8), 1265–1279.