Content uploaded by Oliver Lindemann
Author content
All content in this area was uploaded by Oliver Lindemann
Content may be subject to copyright.
Short article
Motor simulation in verbal knowledge acquisition
Markus Paulus, Oliver Lindemann, and Harold Bekkering
Donders Institute for Brain, Cognition and Behaviour, Radboud University Nijmegen, Nijmegen, The Netherlands
Recent research highlights the importance of motor processes for a wide range of cognitive functions
such as object perception and language comprehension. It is unclear, however, whether the involve-
ment of the motor system goes beyond the processing of information that is gathered through active
action experiences and affects also the representation of knowledge acquired through verbal learning.
We tested this prediction by varying the presence of motor interference (i.e., squeezing a ball vs.
oddball detection task) while participants verbally acquired functional object knowledge and examined
the effects on a subsequent object detection task. Results revealed that learning of functional object
knowledge was only impaired when participants performed an effector-specific motor task while train-
ing. The present finding of an effector-specific motor interference effect on object learning demon-
strates the crucial role of the motor system in the acquisition of novel object knowledge and
provides support for an embodied account to perception and cognition.
Keywords: Embodied cognition; Tool use; Semantic learning; Object perception; Implicit memory.
Imagine yourself ambling through an archaeo-
logical museum and observing the exhibits of
objects from the ancient empires. Some of the
tools used in these times seem very unfamiliar to
you. Fortunately, although you will never experi-
ence their function through your own actions,
you can make sense of these objects through
reading the explanations on the information
panel. As this example illustrates, knowledge
about the functional use of objects can be acquired
even without handling an object. But how do we
acquire functional object knowledge that is not
based on direct sensorimotor experiences?
Developmental research has accumulated
evidence demonstrating that action knowledge
about tools is acquired through motor experiences
(Barrett, Davis, & Needham, 2007) or the obser-
vations of others’ actions (Elsner & Pauen,
2007). These two learning mechanisms indicate
that functional object knowledge goes beyond a
direct association between visual object features
and afforded actions (Tucker & Ellis, 1998; cf.
Gibson, 1979). In the same vein, recent studies
demonstrate that participants are slower to ident-
ify an object depicted in a position that deviates
from its actual correct use than an object depicted
Correspondence should be addressed to Markus Paulus, Donders Institute for Brain, Cognition and Behaviour, Radboud
University Nijmegen, P.O. Box 9104, 6500 HE Nijmegen, The Netherlands. E-mail: m.paulus@donders.ru.nl
We thank Sabine Hunnius for useful discussions of this project as well as Terry Eskenazi for comments on an earlier draft of this
manuscript. The present study was supported by a VICI Grant (453– 05– 001) from the Dutch Organization for Scientific Research
(NWO) and the ICIS project (BSIK03024) sponsored by the Dutch Ministry of Economic Affairs.
#2009 The Experimental Psychology Society 1
http://www.psypress.com/qjep DOI:10.1080/17470210903108405
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY
iFirst, 1–8
Downloaded By: [Radboud Universiteit Nijmegen] At: 18:41 20 September 2009
in a position that is appropriate with respect to the
experienced function (van Elk, Paulus, Pfeiffer,
van Schie, & Bekkering, 2009; van Elk, van
Schie, & Bekkering, 2008). However, it has been
recently argued that object knowledge relies also
on language-related representations and semantic
information processes (e.g., Creem & Proffitt,
2001). Evidence for this notion comes from be-
havioural (Lindemann, Stenneken, van Schie, &
Bekkering, 2006) and neuroimaging studies
(Canessa et al., 2008) showing that semantic infor-
mation concerning the use of objects is activated
during the performance and observation of object-
related actions. For example, Creem and Proffitt
(2001) reported that participants grasped familiar
household tools more frequently inappropriately
with respect to their function when the motor
task was paired with a concurrent semantic task
but not when it was paired with a visuospatial
task suggesting that semantic processing is required
when grasping a tool appropriately for its use.
Several theorists in the field of cognitive psy-
chology have proposed that knowledge about
actions and objects is bodily grounded in sensori-
motor experiences (for reviews, see Barsalou,
2008; Fischer & Zwaan, 2008). Following these
so-called embodied cognition approaches, it is
assumed that the processing of knowledge of func-
tional objects consists in a covert simulation of
associated motor programmes and a reenactment
of the objects’ functional use. Accordingly, neuroi-
maging studies have shown an activation of motor
areas during observation of tools (e.g., Chao &
Martin, 2000). Furthermore, evidence has been
provided that this motor activation during
passive observation of objects is based on one’s
own action experiences with these objects
(Kiefer, Sim, Liebich, Hauk, & Tanaka, 2007).
However, as illustrated by the museum anecdote
above, people can acquire functional knowledge
about an object without having any actual motor
experiences with that object. According to the
view of embodied cognition, also such verbally
acquired action knowledge should be based on
mental simulations of the actual object use.
The present study aimed to test this prediction
derived from the embodied cognition account and
investigated whether the verbal acquisition of
functional object knowledge involves simulation
within the motor system. To do so, we manipu-
lated the presence of motor interference (i.e., an
alternating squeezing of soft balls with the
hands; Witt & Proffitt, 2008) while participants
verbally learned the functions of unknown
objects (i.e., through sentences describing the
object function). If motor simulation mediates
the acquisition of object knowledge, it can be
expected that learning performances are impaired
for participants performing a secondary motor
task as compared to participants conducting a
task without motor demands or no secondary
task. In other words, we hypothesized that a sim-
ultaneously performed motor task affects the par-
ticipants’ capability to simulate the motor action
associated with the object, which should in turn
result in an impaired acquisition of functional
object knowledge. Since neuroimaging studies of
language processing have shown a somatotopically
organized pattern of activation in premotor cortex
for words denoting actions that are related to
different body parts (e.g., Hauk, Johnsrunde, &
Pulvermu¨ller, 2004; Rueschemeyer, Brass, &
Friederici, 2007), one might speculate that motor
simulation while verbal learning is also effector-
specific in nature. To investigate whether the
acquisition of manual functional object knowledge
is differently affected by a motor interference of
another effector than the hand, we introduced an
additional condition, in which participants
performed as a dual task alternating movements
with the feet. To test the functional object knowl-
edge acquired in the learning phase, we used an
object-detecting task that has been shown to be
sensitive to functional object knowledge (van Elk
et al., 2008) and contextual action cues (Fischer,
Prinz, & Lotz, 2008).
Method
Participants
A total of 64 students of the Radboud University
Nijmegen (19–39 years) participated in the experi-
ment in return for 8 euros or course credits.
2THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 0000, 00 (0)
PAULUS, LINDEMANN, BEKKERING
Downloaded By: [Radboud Universiteit Nijmegen] At: 18:41 20 September 2009
Set-up and stimuli
Four novel objects without a predefined function
were constructed (see Figure 1A) and divided into
two object sets. Each object set consisted of one
object that was associated with the action smelling
(smell-object) and one that was associated with
the action hearing (hear-object) so that both
actions were represented in each object set.
Photographs of the objects served as stimuli in the
training phase (object picture) and as primes in
the object recognition task of the test phase. As
target stimuli for the object recognition task, we
used photographs of a person using or holding
the objects in different way (action pictures; see
Figure 1B). In order to reduce stimulus– response
automaticity, we used two different action pictures
for each object use. To be precise, for each object
four different action pictures were taken, in which
the correctness of the object use was systematically
varied. Two action pictures showed a particular
object used correctly with respect to the previously
learned function (correct action; e.g., smell-object
at the nose). The other two pictures depicted an
incorrect object use (incorrect action; e.g., smell-
object at the cheek). Just as the correct action of
every object had a specific position on the person’s
face (e.g., hear-object at ear), the incorrect action
of every object also had a specific position, which
was different for each object and never involved
the nose or the ears. All photographs sustained a
viewing distance of 80 cm and a visual angle of
13 13 degrees.
Figure 1. Part A shows the object pictures used in the experiment. Part B gives an example of the action pictures used in the experiment. In the
example the “smell”-object of Object Set 1 is presented at a correct and an incorrect position regarding the function of the object.
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 0000, 00 (0) 3
MOTOR SIMULATION IN VERBAL KNOWLEDGE ACQUISITION
Downloaded By: [Radboud Universiteit Nijmegen] At: 18:41 20 September 2009
Procedure
Training phase. Participants were indicated to learn
verbally the functional use of one hear-object and
one smell-object (trained objects). No function
was associated with the remaining two objects
(untrained objects). At the beginning of each
trial, participants pressed a button. Afterwards a
fixation cross was presented for 500 ms, followed
by an object picture. If the depicted object was
functional, participants were instructed to release
the button and repeat a standardized sentence
describing the function of the object (“With this
object you can smell [hear] something”). If the
depicted object was not associated with a
functional use, no response was required, and the
next trial was initiated after 3 seconds. The
object picture disappeared after participants fin-
ished the sentence and pressed again the button
to initiate the next trial.
Importantly, participants were randomly
assigned to four different training conditions. In
each condition participants performed the verbal
learning task. In the no interference condition, par-
ticipants merely placed their hands in front of the
button box. In the hand interference condition,
participants performed a hand-related motor task
during sentence articulation. Specifically, they
were instructed to grasp with each hand a squeez-
able foam rubber ball, hold the forearms upwards,
and squeeze the foam balls alternately in the right
and left hand during sentence articulation. In the
foot interference condition participants performed
foot movements as a dual motor task and
squeezed, analogous to the hand interference con-
dition, alternately two rubber balls that were
placed underneath their feet. In the attentional
interference condition, participants performed sim-
ultaneously an auditory oddball target detection
task. That is, during the whole training session
beep tones (1,500 Hz lasting for 5 ms) were pre-
sented alternating at the left and right side. With
a likelihood of 10% the frequency and duration
deviated from the other sounds (i.e., oddball
target; 440 Hz lasting for 250 ms). Participants
had to remember the location of the last oddball
target, because they were occasionally asked to
indicate this by a left/right keypress response.
Test phase. The test phase comprised an object
recognition task similar to the task used by van
Elk et al. (2008). Each trial started with a fixation
cross for 500 ms, followed by a 1,000 ms presen-
tation of an object picture. Another fixation cross
appeared for 1,000 ms and was followed by a
picture of a person using the object. Participants
were required to signal as fast as possible
whether the object in the action picture was the
same as that presented in the first picture or not.
The matching of the object and action pictures
was indicated by a left/right button press response.
The picture disappeared, and the next trial started
immediately after the response was finished.
Design
The four different training conditions (no interfer-
ence, hand interference, attentional interference,
foot interference) were randomly assigned to the
participants. To prevent participants from getting
familiarized with the pictures depicting the incor-
rect use of the object during the test phase, three
training and test phases were presented in an alter-
nating fashion. During each training phase, each
object picture was presented 12 times, resulting
in 24 trials with and 24 trials without sentence
articulation. Each test phase comprised 96 target
trials consisting of the four objects (two trained,
two untrained) each used in two different ways
(twice in a correct way and twice in an incorrect
way). In these target trials the object in the
action picture was the same as that presented in
the first picture, and a “yes”-response was required.
Additionally 48 catch-trials were included where
the object in the action picture was different
from the object in the first picture, and a “no”-
response was required. The training of the two
different object sets was counterbalanced between
participants.
Data analysis
Reaction times (RTs) were measured relative to onset
of the action picture. Trials with incorrect responses,
trials with RTs deviating more than two standard
deviations from the mean RT, and the first three
trials of the first block (practice trials) were excluded
from the subsequent analyses. Trained and untrained
4THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 0000, 00 (0)
PAULUS, LINDEMANN, BEKKERING
Downloaded By: [Radboud Universiteit Nijmegen] At: 18:41 20 September 2009
objects were analysed separately using a two-way
analysis of variance (ANOVA) with the within-
subjects factor object use (correct, incorrect) and the
between-subjects factor training condition (no inter-
ference, hand interference, attentional interference,
foot interference).
Results
Participants incorrectly responded to action pic-
tures in less than 1% of the trials. No difference
was found in the error rates between the four train-
ing conditions (F,1).
The RT analysis for the untrained objects
revealed only a main effect of training condition,
F(3, 60) ¼3.15, p,.05,
h
p
2
¼0.17. Post hoc
comparisons revealed that participants in the foot
interference condition performed faster than par-
ticipants in the no interference and the attentional
interference condition, both ps,.01.
The ANOVA for trained objects revealed a
main effect of object use, F(1, 60) ¼44.15,
p,.001,
h
p
2
¼0.42, reflecting faster responses to
pictures depicting an correct object use (514 ms)
than to those depicting an incorrect object use
(550 ms). Also the factor training condition
reached significance, F(3, 60) ¼2.96, p,.05,
h
p
2
¼0.13. Post hoc comparisons revealed faster
responses in the foot interference (475 ms)
condition than in the no interference condition
(538 ms), t(30) ¼–2.81, p,.01, and the
attentional interference condition (549 ms),
t(30) ¼–2.37, p,.05. Most importantly, the
object use effect was modulated by the different
training conditions as indicated by an significant
interaction between the factors object use and
training condition, F(3, 60) ¼3.11, p,.05,
h
p
2
¼0.14 (see Table 1).
To explore the observed interaction in greater
detail, we computed for each participant the learning
effect defined as average RT difference between
trials with correct and incorrect object use (see
Figure 2). Interestingly, we observed substantial
learning effects for the conditions no interference,
one-sample t(15) ¼3.81,p,.01, attentional inter-
ference, one-sample t(15) ¼3.46,p,.01, and foot
interference, one-sample t(15) ¼4.29,p,.01.
However, there was no learning effect for the hand
interference condition, one-sample t(15) ¼1.83,
p..09. Pairwise ttests revealed furthermore that
average RT difference in the condition hand
interference (11 ms) was significantly smaller than
that in the conditions no interference (56 ms),
t(30) ¼2.83, p,.01, attentional interference
Table 1. Mean reaction times and standard errors to identify trained and untrained objects as a function of the different training conditions
and the correctness of the depicted actions
Trained object Untrained objects
Training condition Correct action Incorrect action Correct action Incorrect action
No interference 510 (19) 565 (19) 561 (18) 571 (18)
Hand interference 504 (16) 515 (13) 538 (14) 536 (13)
Attentional interference 530 (28) 569 (29) 569 (29) 566 (27)
Foot interference 458 (15) 492 (14) 490 (17) 497 (16)
Note: Reaction times in ms; standard errors in parentheses.
Figure 2. Mean reaction time differences between incorrect and
correct actions (dRT) as a function of the different training
conditions (no interference, hand interference, attentional
interference, and foot interference). Error bars indicate the
standard errors.
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 0000, 00 (0) 5
MOTOR SIMULATION IN VERBAL KNOWLEDGE ACQUISITION
Downloaded By: [Radboud Universiteit Nijmegen] At: 18:41 20 September 2009
(39 ms), t(30) ¼–2.20, p,.05, and foot interfer-
ence (33 ms), t(30) ¼–2.28, p,.05, which were
not different from each other, all ps..19.
Discussion
The present study aimed to examine the role of
motor simulation for the acquisition of functional
object knowledge and demonstrates a selective
impairment of the verbal learning of object func-
tions if it is accompanied by an execution of a
manual motor action (i.e., squeezing a ball). This
finding of a selective motor interference effect on
object learning is in line with the idea that covert
motor simulations support the acquisition of
functional object knowledge.
Previous research (van Elk et al., 2008) has
shown that performances in object detection
tasks reflect participants’ functional object knowl-
edge by showing that objects presented at the
associated action goal location are recognized
faster than objects at another location (e.g., cup
at eye). The calculated difference between the
object detection times toward correct and incorrect
action in the test phase could be consequently
interpreted as an indicator for the functional
object knowledge that participants acquired
during the learning phase. Accordingly, partici-
pants in the no interference condition showed a
substantial learning effect indicated by the facili-
tated detections of objects presented at its correct
action goal location. Interestingly, this learning
effect vanished if participants were required to
perform a secondary manual motor task during
the learning phase (hand interference condition).
Importantly, we can exclude that the interference
effect reflects a general deficit of cognitive resources
or an attention effect because learning was unaf-
fected by the oddball detection task (attentional
interference condition) as well as by the foot
movement task (foot interference condition).
Furthermore, participants had no difficulties
remembering the correct sentences in the training
phase, rendering it unlikely that the secondary
task impaired the verbal performance itself.
Additionally, the learning effect cannot be attri-
buted to any perceptual differences between the
pictures because the analysis for the reaction times
to the untrained objects did not reveal differences
in the recognition time between action pictures
containing correct and incorrect actions.
The results demonstrate moreover that the
acquisition of manual object knowledge was selec-
tively impaired as the consequence of the concur-
rent manual action but not if concurrent actions
with the feet were performed. This finding
suggests that covert motor simulations are effec-
tor-specific and is thus in line with neuroimaging
studies showing effector-specific cortical acti-
vations while action-word reading (e.g., Hauk
et al., 2004; Rueschemeyer et al., 2007). Based
on this literature, one might speculate that covert
simulations of hand-related motor actions were
selectively impaired while ball squeezing (hand
interference condition) as the result of an effec-
tor-specific activation of motor areas in the brain.
The present finding of a motor interference
effect on the acquisition of functional object
knowledge goes beyond previous research that
claimed that object recognition relies on motor
knowledge about the use of an object (Canessa
et al., 2008; Chao & Martin, 2000). This claim
was indirectly supported by a recent study in
which participants were trained with novel
objects (Kiefer et al., 2007). Participants either
had to make an action pantomime towards the
object displaying its use or had to point to it.
Interestingly, only the pantomime group showed
activations in motor areas when confronted again
with the objects showing the influence of action
knowledge on object processing. It is important
to note that participants in our study did not
acquire knowledge about the functions of novel
objects through own action experiences. Despite
the fact that learning occurred purely verbally
and without active interactions with the object,
we observed that functional object knowledge
was selectively impaired by a concurrent motor
task. Our study therefore suggests that motor pro-
cesses also underlie the verbal acquisition of object
knowledge, which is not based on own action
experiences.
However, whereas it is clear that a concurrent
motor task impairs the acquisition of functional
6THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 0000, 00 (0)
PAULUS, LINDEMANN, BEKKERING
Downloaded By: [Radboud Universiteit Nijmegen] At: 18:41 20 September 2009
object knowledge it remains unclear whether the
effect of this knowledge on the perceptual task is
based on a faster detection of compatible trials or
on a cognitive interference in incompatible trials
due to the overall reaction time differences
between the four conditions. Further research
involving a neutral baseline is needed to address
this question and to differentiate between both
possibilities.
The present results not only add to our under-
standing of the representation of functional object
knowledge but may also have implications for an
embodied theory of language processing. Studies
provided evidence that language processing auto-
matically activates effector- and modality-specific
subsystems (for an overview, see Pulvermu¨ ller,
2005) and that it behaviourally interferes with
perceptual and motor processes (Glenberg &
Kaschak, 2002; Zwaan, Stanfield, & Yaxley,
2002) suggesting that perceptuo-motor processes
contribute to the understanding of language.
However, it is unclear whether the activation of
motor representations is indeed necessary for
language comprehension or if the activation of
the motor system is merely a by-product of an
amodal information processing (Fischer &
Zwaan, 2008). If perceptuomotor simulations are
indeed necessary for language comprehension we
would expect that the verbal acquisition of novel
object knowledge should be affected by a concur-
rent motor task. Our finding that an occupied
manual motor system affects selectively the
verbal acquisition of new functional object knowl-
edge could be thus interpreted in accord with a
strong embodied approach. However, future
research is needed to test this speculation directly
and demonstrate that our finding of an effector-
specific motor interference effect on semantic pro-
cessing while object learning can be generalized to
other language-related processes.
In summary, the present study demonstrates
that verbal acquisition of novel functional object
knowledge is selectively impaired while performing
a concurrent manual motor task. Our finding of an
effector-specific motor interference effect on object
learning provides evidence for the crucial role of
the motor system in knowledge acquisition and
for the claim that the processing of knowledge
about functional objects consists in a covert
simulation of associated motor programmes.
Original manuscript received 13 March 2009
Accepted 7 April 2009
First published online day month year
REFERENCES
Barrett, T. M., Davis, E. F., & Needham, A. (2007).
Learning about tools in infancy. Developmental
Psychology,43(2), 352–368.
Barsalou, L. W. (2008). Grounded cognition. Annual
Review of Psychology,59, 617–645.
Canessa, N., Borgo, F., Cappa, S. F., Perani, D., Falini,
A., Buccino, G., et al. (2008). The different neural
correlates of action and functional knowledge in
semantic memory: An fMRI study. Cerebral Cortex,
18, 740–751.
Chao, L., & Martin, A. (2000). Representation of
manipulable man-made objects in the dorsal
stream. NeuroImage,12, 478–484.
Creem, S. H., & Proffitt, D. R. (2001). Grasping objects
by their handles: A necessary interaction between cog-
nition and action. Journal of Experimental Psychology:
Human Perception and Performance,27, 218–228.
Elsner, B., & Pauen, S. (2007). Social learning of artifact
function in 12- and 15-month-olds. European
Journal of Developmental Psychology,4, 80–99.
Fischer, M. H., Prinz, J., & Lotz, K. (2008). Grasp
cueing shows obligatory attention to action goals.
Quarterly Journal of Experimental Psychology,61,
860–868.
Fischer, M. H., & Zwaan, R. A. (2008). Embodied
language: A review of the role of the motor system
in language comprehension. Quarterly Journal of
Experimental Psychology,61, 825–850.
Gibson, J. J. (1979). The ecological approach to visual
perception. Boston: Houghton Mifflin.
Glenberg, A. M., & Kaschak, M. P. (2002). Grounding
language in action. Psychonomic Bulletin & Review,
9, 558–565.
Hauk, O., Johnsrunde, I., & Pulvermu¨ ller, F. (2004).
Somatotopic representation of action words in
human motor and premotor cortex. Neuron,41(2),
301–307.
Kiefer, M., Sim, E.-J., Liebich, S., Hauk, O., & Tanaka,
J. (2007). Experience-dependent plasticity of
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 0000, 00 (0) 7
MOTOR SIMULATION IN VERBAL KNOWLEDGE ACQUISITION
Downloaded By: [Radboud Universiteit Nijmegen] At: 18:41 20 September 2009
conceptual representations in human sensory-motor
areas. Journal of Cognitive Neuroscience,19, 525–542.
Lindemann, O., Stenneken, P., van Schie, H. T., &
Bekkering, H. (2006). Semantic activation in
action planning. Journal of Experimental Psychology:
Human Perception and Performance,32, 633–643.
Pulvermu¨ ller, F. (2005). Brain mechanisms linking
language and action. Nature Review Neuroscience,6,
576–582.
Rueschemeyer, S. A., Brass, M., & Friederici, A. D.
(2007). Comprehending prehending: Neural corre-
lates of processing verbs with motor stems. Journal
of Cognitive Neuroscience,19(5), 855–865.
Tucker, M., & Ellis, R. (1998). On the relations
between seen objects and components of potential
actions. Journal of Experimental Psychology: Human
Perception and Performance,24, 830–846.
van Elk, M., Paulus, M., Pfeiffer, C., van Schie, H. T.,
& Bekkering, H. (2009). The role of object’s action
effects in the acquisition of novel action semantics.
Manuscript submitted for publication.
van Elk, M., van Schie, H. T., & Bekkering, H. (2008).
Conceptual knowledge for understanding other’s
actions is organized primarily around action goals.
Experimental Brain Research,189, 99–107.
Witt, J. K., & Proffitt, D. R. (2008). Action-specific
influences on distance perception: A role for
motor simulation. Journal of Experimental
Psychology: Human Perception and Performance,34,
1479–1492.
Zwaan,R.A.,Stanfield,R.A.,&Yaxley,R.H.
(2002). Language comprehenders mentally
represent the shapes of objects. Psychological
Science,13, 168 – 171.
8THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 0000, 00 (0)
PAULUS, LINDEMANN, BEKKERING
Downloaded By: [Radboud Universiteit Nijmegen] At: 18:41 20 September 2009
