The analgesic effect of crossing the arms
, D.M.E. Torta
, G.L. Moseley
, G.D. Iannetti
Department of Psychology, University of Milano-Bicocca, Italy
Department of Neuroscience, Physiology and Pharmacology, University College London, UK
Department of Psychology, University of Turin, Italy
The Sansom Institute for Health Research, University of South Australia, Adelaide, and Neuroscience Research Australia, Sydney, Australia
Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.
Received 24 May 2010
Received in revised form 22 December 2010
Accepted 10 February 2011
Event-related potentials (ERPs)
The ability to determine precisely the location of sensory stimuli is fundamental to how we interact with
the world; indeed, to our survival. Crossing the hands over the body midline impairs this ability to local-
ize tactile stimuli. We hypothesized that crossing the arms would modulate the intensity of pain evoked
by noxious stimulation of the hand. In two separate experiments, we show (1) that the intensity of both
laser-evoked painful sensations and electrically-evoked nonpainful sensations were decreased when the
arms were crossed over the midline, and (2) that these effects were associated with changes in the mul-
timodal cortical processing of somatosensory information. Critically, there was no change in the somato-
sensory-speciﬁc cortical processing of somatosensory information. Besides studies showing relief of
phantom limb pain using mirrors, this is the ﬁrst evidence that impeding the processes by which the
brain localises a noxious stimulus can reduce pain, and that this effect reﬂects modulation of multimodal
neural activities. By showing that the neural mechanisms by which pain emerges from nociception rep-
resent a possible target for analgesia, we raise the possibility of novel approaches to the treatment of
painful clinical conditions.
Ó2011 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
Pain reduces tissue damage by motivating escape . In order to
be fully effective for survival, the ability to localize where a noxious
stimulus occurs must be as accurate as possible . Furthermore,
the processing of the spatial information of a sensory input is one
of the important requisites for it reaching awareness, as impeding
the processing of a stimulus’ location often impedes also the percep-
tion of that stimulus [17,18]. This raises the possibility that disrupt-
ing the processes by which the location of a sensory stimulus is
determined will reduce the perceived intensity of that stimulus.
Remarkably, disrupting the very processes by which nociceptive
input emerges into awareness  by acting upon its correct spa-
tial localization has not, until now, been targeted as a method of
reducing pain in healthy volunteers, although the mislocalization
of noxious stimuli by a mirror has been shown to result in reduc-
tion of pain in patients with phantom limb . In contrast, there
have been successful attempts to impede awareness of nociceptive
stimuli by acting upon a person’s state of consciousness or atten-
The ability to localise tactile inputs is impaired when hands are
crossed over the body midline . For example, when 2 sequential
stimuli are presented, one on each hand, crossing the hands over
the body midline reduces our ability to determine which hand
was stimulated ﬁrst [15,53]. This ‘‘crossed-hands deﬁcit’’ is
thought to occur because of a mismatch between the location of
the stimulus within an anatomical (or somatotopical) frame of ref-
erence and the location of the stimulus within a space-based frame
of reference [4,16,53]. Indeed, to localize correctly sensory inputs
in the environment, most of the somatosensory experience must
be referred to spatial locations deﬁned according to nonsomato-
topical frames of reference [20,42,45]. It is thought that integration
of information between these 2 frames of reference probably oc-
curs in multimodal brain areas, that is, areas responding to stimuli
of different sensory modalities .
So far, crossed-hands studies have not explored the intensity of
perception or the neural activity elicited by somatosensory stimuli
delivered to the hands. Given that (1) crossing the hands over the
body midline impairs the ability to localise tactile stimuli ; (2)
localization of tactile stimuli is an important requisite for awareness
(eg, ); and (3) multimodal brain areas (eg, associational and lim-
bic areas that respond to stimuli belonging to different sensory
modalities [2,32]) have been hypothesized to play a more important
role in awareness of both tactile  and noxious  stimuli than
somatosensory-speciﬁc brain areas do, we hypothesized that
0304-3959/$36.00 Ó2011 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
Corresponding author. Address: Department of Neuroscience, Physiology and
Pharmacology, University College London, Medical Sciences Building, Gower Street,
London WC1E 6BT, UK. Tel.: +44 (0) 20 7679 3759.
E-mail address: firstname.lastname@example.org (G.D. Iannetti).
152 (2011) 1418–1423
crossing the arms would impede multimodal processing of somato-
sensory stimuli delivered to the hands, and thereby decrease the
intensity of both painful and tactile sensations. As multimodal pro-
cessing of somatosensory stimuli is reﬂected in the late components
of the event-related potentials (ERPs) elicited by somatosensory
stimuli , we would expect a stronger modulation of late rather
than early ERP components when the hands are crossed over the
To test these 2 hypotheses, we investigated the intensity of the
sensations (Experiment 1) and the brain activity (Experiment 2)
evoked by nonnociceptive electrical stimuli and nociceptive laser
stimuli delivered to the hands, while the subject’s arms were either
crossed or not crossed over the body midline.
2. Material and methods
Eight right-handed healthy subjects (4 women, mean age
28 ± 4.7 years) took part in Experiment 1. Twelve right-handed
healthy subjects (6 women, mean age 31 ± 10.6 years) took part in
Experiment 2. All participants gave their written informed consent.
This study conformed to the standards required by the Declaration of
Helsinki and was approved by the institutional ethics committee.
Participants were comfortably seated in a silent, temperature-
controlled room, and wore protective laser-proof goggles. Earplugs
and headphones were also worn in order to remove any possibly
auditory cue arising from the operation of the devices. Two large
wooden screens placed in front of the participants occluded their
view of both their arms and the experimenter. A ﬁxation point was
placed 30 cm in front of, and 30°below their eyes. Before the begin-
ning of the experiments participants were familiarised with the sen-
sory stimulation (5–10 stimuli of each modality). During the
experiment the participants were asked to keep their arms either un-
crossed or crossed over the midline. The distance between the hands
was the same (40 cm) in the 2 conditions. Crossed and uncrossed
blocks were run separately (see below for further details).
2.2.1. Nonnociceptive somatosensory stimuli
Nonnociceptive somatosensory stimuli consisted of constant
current square-wave pulses (1-ms duration; DS7A, Digitimer, Hert-
fordshire, UK) delivered through 2 electrodes (0.5 cm diameter,
2 cm interelectrode distance) placed at the wrist, over the superﬁ-
cial radial nerve. Stimulus intensity was adjusted to elicit a clear,
nonpainful paresthesia. Electrical stimuli were always delivered
at an intensity that was above the threshold of Abﬁbres (which
convey nonnociceptive tactile information) but below the thresh-
old of nociceptive Adand C ﬁbres . In all experiments, electrical
stimuli were never reported as painful.
2.2.2. Nociceptive somatosensory stimuli
Nociceptive somatosensory stimuli consisted of 4-ms pulses of
radiant heat generated by an infrared neodymium yttrium alumin-
ium perovskite (Nd:YAP) laser (wavelength 1.34
m, El.En. Group,
Florence, Italy). Beam diameter at target site was 8 mm. Laser
stimuli were delivered to the sensory territory of the superﬁcial
radial nerve. Stimulus energy was initially adjusted to elicit a clear
painful pinprick sensation, related to the selective activation of Ad
skin nociceptors, thus solving the previously problematic issue of
the co-activation of nonnociceptive afferents [7,10]. To prevent
nociceptor fatigue or sensitization, the laser target was displaced
after each pulse.
Both experiments involved 8 blocks of 30 stimuli each (inter-
stimulus interval randomised between 8 and 12 seconds). Fifteen
stimuli were delivered to each hand in pseudorandom counterbal-
anced order. The maximum number of consecutive stimuli deliv-
ered to the same hand was 3. Both stimulus modality
(nonnociceptive and nociceptive) and hand position (crossed or
uncrossed) were the same within each block. There were 2 blocks
each of ‘‘nonnociceptive crossed,’’ ‘‘nonnociceptive uncrossed,’’
‘‘nociceptive crossed,’’ and ‘‘nociceptive uncrossed.’’ The order of
blocks was balanced across participants.
2.2.3. Experiment 1
In Experiment 1, 3 energies of both nonnoxious and noxious stim-
ulations were determined using two, 0–100 numerical rating scales,
one for the sensation elicited by laser stimuli and one for the sensa-
tion elicited by electrical stimuli. The anchors of the scale used to
rate the intensity of perception elicited by laser stimuli were ‘‘no pin-
prick sensation’’ (0) and ‘‘the strongest pinprick sensation imagin-
able’’ (100). When rating laser stimuli, participants were explicitly
instructed to rate 0 in response to a nonpricking (eg, nonpainful
warm) sensation. The anchors of the scale used to rate the intensity
of perception elicited by electrical stimuli were ‘‘no electrical sensa-
tion’’ (0) and ‘‘the strongest electrical sensation imaginable’’ (100).
‘‘Low,’’ ‘‘medium,’’ and ‘‘high’’ energies, corresponding to ratings of
30, 50, and 70, were 4.1 ± 1.7 mA, 6.8 ± 2.3 mA, and 10.2 ± 3.4 mA
for electrical and 2 J, 2.5 J, and 3 J for laser stimulation, respectively.
At these energies, Nd: YAP laser pulses consistently elicit painful
sensations (eg, ). Ten stimuli of each energy were delivered in
pseudorandom counterbalanced order. The maximum number of
consecutive stimuli of the same energy was 3. Participants rated
the perceived intensity of each stimulus.
2.2.4. Experiment 2
In Experiment 2, one stimulus energy (11.8 ± 4.0 mA for electri-
cal and 3.0 ± 0.5 J for laser) was used. The electroencephalogram
was recorded using 21 scalp electrodes placed according to the
International 10-20 system. The nose was used as common refer-
ence. Signals were ampliﬁed and digitized (sampling rate
1024 Hz; resolution = 0.195
; System Plus, Micromed,
Italy), segmented into 1-second epochs (200 to +800 ms relative
to stimulus onset) and band-pass ﬁltered (1–30 Hz). After baseline
correction (reference interval 200 to 0 ms), artefacts due to eye
blinks or movements were subtracted using a validated method
. Epochs exceeding ± 100
V amplitude (ie, likely to be con-
taminated by artefact) were rejected. They were 0.8 ± 1.6% of the
total number of epochs. Separate average ERP waveforms time-
locked to stimulus onset were computed for each participant,
according to stimulus modality (nonnociceptive or nociceptive)
and hand position (crossed or uncrossed). Baseline-to-peak ampli-
tude of the N1 wave, which represents an early stage of sensory
processing and reﬂects the somatosensory-speciﬁc afferent input
and its somatotopical arrangement, and peak-to-peak amplitude
of the N2-P2 wave, which represents a later stage of processing
and mostly reﬂects multimodal neural activities [27,40,50], were
measured. The labels N1 and N2-P2 for the ERPs elicited by both
nonnociceptive and nociceptive stimuli were chosen according to
Treede et al . Analyses were conducted using Letswave 
and Matlab (MathWorks, Natick, MA, USA).
3.1. Experiment 1
For each of the 3 energies used, laser stimuli elicited a clear pin-
prick pain in all participants, related to the activation of Adﬁbres
A. Gallace et al. / PAIN
152 (2011) 1418–1423 1419
. Ratings of perceived intensity are reported in Table 1. A 3-way
analysis of variance (‘‘stimulus modality’’ [nonnociceptive or noci-
ceptive] ‘‘arm position’’ [crossed or uncrossed] ‘‘stimulus en-
ergy’’ [low, medium, or high]) revealed that crossing the arms
reduced the intensity of the sensation evoked by the stimuli,
regardless of their sensory modality and of the energy of the
applied stimulus (main effect of ‘‘arm position’’: F(1, 6) = 7.54,
P= 0.03; ‘‘stimulus modality’’ ‘‘arm position’’ interaction:
F(1, 6) = 0.55, P= 0.48; ‘‘stimulus energy’’ ‘‘arm position’’ interac-
tion: F(1, 6) = 0.07, P= 0.91). That is, crossing the arms decreased
the intensity of the sensations elicited by both nonnoxious and
noxious stimuli (Fig. 1).
3.2. Experiment 2
Amplitudes of ERPs are reported in Table 2. A 2-way analysis of
variance (‘‘stimulus modality’’ [nonnociceptive or nociceptive]
‘‘arm position’’ [crossed or uncrossed]) revealed that crossing the
arms over the midline reduced the amplitude of the N2-P2 wave,
regardless of stimulus modality (main effect of ‘‘arm position’’:
F(1, 11) = 9.27, P= 0.01; ‘‘stimulus modality’’ ‘‘arm position’’
interaction: F(1, 11) = 0.04, P= 0.81). In contrast, crossing the arms
did not reduce the amplitude of the N1 wave for either stimulus
modality (main effect of ‘‘arm position’’: F(1, 11) = 1.43, P= 0.26;
‘‘stimulus modality’’ ‘‘arm position’’ interaction: F(1, 11) = 0.16,
P= 0.68), which shows no effect of crossing the arms on somato-
sensory-speciﬁc processing (Fig. 2).
We hypothesized that crossing the arms would impede multi-
modal processing of somatosensory stimuli delivered to the hands,
and thereby decrease pain and touch. Our results clearly uphold
this hypothesis – crossing the arms reduced the perceived inten-
sity of both laser-evoked painful sensations and electrically evoked
nonpainful sensations, as evidenced by the behavioural data
(Experiment 1), and selectively disrupted multimodal processing
of both nonnociceptive and nociceptive somatosensory stimuli,
as evidenced by the decreased amplitude of the N2-P2 wave
(Experiment 2). The clear dissociation between the absence of N1
wave modulation and the presence of N2-P2 wave modulation
indicates a clear effect of crossing the arms on multimodal, but
not somatosensory-speciﬁc, neural processing.
One possible explanation for this effect relies on the cognitive
costs associated with realigning neural representations based on
different spatial frames of reference. When we cross our hands,
the conﬂict between the sensory inputs represented in different
frames of reference requires the brain to realign somatosensory
coordinates to spatial coordinates, which has a cost in terms of pro-
cessing resources . The modulation of the neural components
reﬂected in the N2-P2 peaks is likely to represent the neural corre-
late of this cost, while the reduced intensity of the sensations elic-
ited by both nonnoxious and noxious stimuli represents its
A second possible explanation is related to the fact that, when
the arms are held in an uncommon posture (ie, when they are
crossed), the relevance of the stimuli delivered on the hands might
be reduced. Indeed, it has been repeatedly reported that the atten-
tional context in which the eliciting stimulus is presented alters
the magnitude of the N2 and P2 peaks of somatosensory ERPs, even
if the intensity of the afferent input is constant (eg, [21,27–29]).
It has been suggested that nonnociceptive somatosensory infor-
mation is initially processed in a somatotopic frame of reference,
and needs to be later transformed into a more abstract frame of
reference to become available for conscious processing ([24,53];
though see ). Thus, the crossed-hand deﬁcit has been so far
interpreted in terms of the process of progressive ‘‘recoding’’ of
sensory information throughout different spatial maps .
However, it is well known that the several cortical maps in pri-
mate sensory systems are activated both in series and in parallel,
and that they are heavily interconnected . There is also evi-
dence of parallel processing in the human somatosensory system
(eg, [25,33]). For example, after ischemic injury to one entire pri-
mary somatosensory area, patients can be completely unaware of
tactile stimuli delivered to the contralateral body side, but still able
to point correctly to where they occurred [6,42,51]. This empirical
evidence of a somatosensory equivalent of blindsight [44,52] sug-
gests that spatial information regarding tactile stimuli can be pro-
cessed and integrated with motor commands, without primary
somatosensory cortex involvement, possibly by direct anatomical
connections between the lateral posterior thalamic nuclei and
the posterior parietal cortex [6,18,22,33].
On that basis, an alternative possibility can be suggested: that
tactile stimuli, rather than being sequentially converted from one
frame of reference to another, are always mapped both in a
somatosensory and in a space-based representation, and that this
dual mapping happens before conscious judgments are made.
Obviously, the strength of activation of each map might be modu-
lated by a number of parameters (such as the availability of visual
and proprioceptive information, as well as the immediate
relevance of the task; eg, ). For example, seeing the arms might
enhance the neural representation of sensory stimuli in the space-
based map. Thus, the modulation of the perceived intensity of a
stimulus delivered to the hands while they are crossed over the
midline might not be the consequence of the need to remap the in-
put from the somatosensory to the spaced-based map, but of a lack
of correspondence between the ‘‘expected’’ neural activities elic-
ited by the stimulus in these 2 maps.
In everyday life, the right and left hands manipulate objects and
are exposed to somatosensory stimuli that are more often present
on the right or on the left side of space, respectively. For example,
stimuli activating mechanoreceptors on the left hand occur much
more often on the left side of the body midline. Thus, the represen-
tation of the hand in the somatotopic map is often activated to-
gether with the representation of the left side of space in the
space-based map (Fig. 3). Consequently, it is likely that, among
the extensive connections between these 2 maps [17,18], those
between the regions more often receiving a sensory input resulting
in a simultaneous pattern of activity (eg, the left hand area in the
somatosensory map and the left side area of the space-based
map) are likely to display increased synaptic strengths. When our
hands are uncrossed, the match between the 2 frames of reference
makes the processing of sensory stimuli delivered to the hands
Experiment 1: behavioural results.
Hands uncrossed Hands crossed
Energy 1 18.4 ± 10 17.2 ± 10
Energy 2 32.8 ± 9 31 ± 9
Energy 3 45.3 ± 13 43.5 ± 12
Energy 1 20.3 ± 8 17.8 ± 8
Energy 2 48.2 ± 9 44.6 ± 9
Energy 3 60.6 ± 8 57.7 ± 9
Values are expressed as mean ± SD. Statistical comparisons are reported in the text.
Values represent intensity of perception according to a 0–100 numerical scale,
where 0 represents ‘‘no sensation’’ and 100 ‘‘the strongest electrical sensation
Values represent intensity of perception according to a 0-100 numerical scale,
where 0 represents ‘‘no sensation’’ and 100 ‘‘the strongest pinprick sensation
1420 A. Gallace et al. / PAIN
152 (2011) 1418–1423
highly effective in enhancing the sensory signal due to the privi-
leged synaptic connections between the corresponding areas of
the 2 maps (Fig. 3, left panel). In contrast, when we hold an uncom-
mon body posture, such as crossing the hands over the midline,
these privileged synaptic connections are not engaged (Fig. 3, right
panel). Therefore, although the correct localisation of stimuli in
space is still possible, the enhancement of the sensory signal is im-
peded, which might result in decreased intensity of perception.
Interestingly, it has been recently shown that performance in tem-
poral order judgment tasks, which are commonly used to investi-
gate the reference frames involved in the localisation of
somatosensory inputs (as well as the temporal aspects of our
awareness), is abnormal when the hands are crossed over the mid-
line , but only in children older than 5 years . This sug-
gests that, ontogenetically, somatosensory events are referred to
nonsomatotopical frames of references only after the development
of space-based maps. Before the development of those maps, our
ability to localise somatosensory stimuli relies completely on ana-
tomical frames of reference.
The magnitude of early components of the response elicited by
somatosensory stimuli (eg, the N1 wave of laser-evoked
Fig. 1. Behavioural results (Experiment 1). The left panel represents the intensity of the sensations evoked by nonnoxious stimuli (left graph) and noxious stimuli (right
graph), which were delivered to the hands while the arms were either crossed (in blue) or uncrossed (in red). Three energies of both types of somatosensory stimuli were
used. Crossing the arms signiﬁcantly reduced the intensity of the sensation evoked by the stimuli, regardless of their sensory modality (main effect of ‘‘arm position’’: P= 0.03;
‘‘stimulus modality’’ ‘‘arm position’’ interaction: P= 0.48).
Experiment 2: event-related potential (ERP) results.
Hands uncrossed Hands crossed
N1 wave 3.8 ± 4.9 4.5 ± 4.8
N2-P2 wave 34.0 ± 14.5 31.5 ± 13.9
N1 wave 4.2 ± 3.6 4.8 ± 2.6
N2-P2 wave 36.7 ± 19.3 33.2 ± 17.5
Values represent peak amplitude in
V (mean ± SD). Statistical comparisons are
reported in the text.
Fig. 2. Electrophysiological results (Experiment 2). (A) Grand-average waveforms showing the N1 and N2-P2 waves of somatosensory-evoked potentials elicited by
nonnociceptive electrical stimuli (SEPs, left panel) and nociceptive laser stimuli (LEPs, right panel) delivered to the hand dorsum while arms were crossed (blue waveforms)
and uncrossed (red waveforms). Crossing the arms signiﬁcantly reduced the peak amplitude of the N2-P2 wave, regardless of stimulus modality (main effect of ‘‘arm
position’’: P= 0.01; ‘‘arm position’’ ‘‘stimulus modality’’ interaction: P= 0.81). In contrast, crossing the arms did not reduce the peak amplitude of the N1 wave for either
stimulus modality (main effect of ‘‘arm position’’: P= 0.26, ‘‘arm position’’ ‘‘stimulus modality’’ interaction: P= 0.68). The clear dissociation between the absence of N1
wave modulation and the presence of a strong N2-P2 wave modulation indicates that the analgesic effect imparted by crossing the arms involves multimodal, but not
somatosensory-speciﬁc, processing. (B) Group-level scalp distribution of SEPs (top panel) and LEPs (bottom panel) elicited by stimulation of the hand dorsum while the arms
were crossed (top row in each panel) and uncrossed (bottom row in each panel). Scalp maps are displayed at 20-ms intervals, from stimulus onset to 500 ms poststimulus.
A. Gallace et al. / PAIN
152 (2011) 1418–1423 1421
potentials), which better reﬂects the magnitude of the ascending
somatosensory volley , was not affected by crossing the hands.
This excludes the possibility that the change in body posture mod-
ulated the magnitude of the afferent somatosensory input. This
ﬁnding is consistent with single-cell recordings in monkeys that
showed that the neural responses in S1 are correlated to the inten-
sity of the applied stimulus but not to its awareness . In hu-
mans, a clear response of primary somatosensory neurons to
tactile stimuli is observed even when the stimuli are below percep-
tual threshold . In contrast, only late components of the
somatosensory-evoked potentials are correlated with stimulus
awareness [27,43,46,48]. Consistent with these observations, the
observed reduction of perceived intensity consequent to crossing
the arms was only reﬂected in the reduced magnitude of the late
components of the response elicited by somatosensory stimuli
(the N2-P2 wave of laser-evoked potentials), which, despite the
difﬁculties related to the limited reliability of source analysis of
EEG data and blind source separation approaches , have been
suggested to be largely explained by neural activities arising from
multimodal cortical areas . Thus, the selective modulation of
the N2-P2 response when the hands are crossed indicates that
the observed analgesic effect is related to a modulation of the
activity in multimodal cortical areas.
Which multimodal brain areas may be involved in this phenom-
enon? In both human and nonhuman primates, the posterior
parietal cortex is important for the integration of spatial informa-
tion coming from different sensory modalities [2,8,11,19,34].In
particular, the ventral aspect of the intraparietal sulcus, which di-
vides the parietal lobe into the superior and the inferior parietal
lobules , contains neurons that encode the information con-
tained in stimuli belonging to different sensory modalities into a
reference frame that can be accessed by all sensory systems .
Part of the human intraparietal sulcus has been shown to play a
pivotal role in the multisensory representation of limb position
[5,31]. Thus, this human area homologous to the ventral aspect
of the intraparietal sulcus in nonhuman primates is also likely to
be related to the modulation of both somatosensory perception
(Fig. 1) and the magnitude of multimodal ERP components (Fig. 2).
This is the ﬁrst evidence that disrupting the processes by which
the brain localises a noxious stimulus reduces the pain evoked by
that stimulus. The magnitude of the effect shown here is too small
to be clinically important, but it reveals for the ﬁrst time that the
mechanisms by which a sensory event emerges into awareness
can modulate pain. This extends a previous result that perceptual
illusions can modulate pain , and raises a new possible expla-
nation for the purported analgesic effect of mirror therapy,
although that mirror analgesia is due to seeing the reﬂected image
is not established, and other explanations are possible . Finally,
the current results lay the platform for future studies that maxi-
mise conﬂict between neural representations of a noxious stimulus
according to somatotopic and space-based frames of reference,
possibly resulting in larger and clinically important analgesic ef-
fects. Perhaps, when we get hurt, we should not only ‘‘rub it better’’
but also cross our arms.
Fig. 3. A putative neurocognitive model supporting the presented ﬁndings. Tactile, nociceptive, and proprioceptive information arising from, for example, the right hand
(black arrows), reaches the corresponding areas in the somatosensory cortices. When the hands are uncrossed (left panel), these inputs also activate multisensory areas
mapping the right side of the external space, with reference to the body midline. Thus, stimulation of each hand results in a match between somatosensory and space-based
representations (thick, green double arrow). When the hands are crossed (right panel), somatosensory information arising from, for example, the right hand (black arrows),
reaches the corresponding areas in the somatosensory cortices but, critically, also those multisensory areas mapping the left side of the external space. Thus, stimulation of
each hand results in a mismatch between somatosensory and space-based representations (thin, green-red double arrow). Note that the neural connections between spatially
corresponding areas of the somatotopic and space-based representations (eg, the somatotopic representation of the right hand and the representation of the right side of the
external space) are stronger than those between areas of the somatotopic and space-based representations that do not correspond (see main text for an explanation of the
reasons underlying this assumption). The mismatch between somatosensory and space-based representations results in a reduced perceived intensity of the delivered
1422 A. Gallace et al. / PAIN
152 (2011) 1418–1423
Conﬂict of interest statement
The authors declare no conﬂict of interest.
A.G. is supported by a MIUR PRIN07 Grant. G.L.M. is supported
by a Senior Research Fellowship from the National Health and
Medical Research Council of Australia. G.D.I. is University Research
Fellow of The Royal Society.
 Aglioti S, Smania N, Peru A. Frames of reference for mapping tactile stimuli in
brain-damaged patients. J Cogn Neurosci 1999;11:67–79.
 Andersen RA. Multimodal integration for the representation of space in the
posterior parietal cortex. Philos Trans R Soc Lond B Biol Sci 1997;352:1421–8.
 Avillac M, Deneve S, Olivier E, Pouget A, Duhamel JR. Reference frames for
representing visual and tactile locations in parietal cortex. Nat Neurosci
 Azanon E, Soto-Faraco S. Changing reference frames during the encoding of
tactile events. Curr Biol 2008;18:1044–9.
 Bolognini N, Maravita A. Proprioceptive alignment of visual and
somatosensory maps in the posterior parietal cortex. Curr Biol
 Brochier T, Habib M, Brouchon M. Covert processing of information in
hemianesthesia: a case report. Cortex 1994;30:135–44.
 Bromm B, Treede RD. Nerve ﬁbre discharges, cerebral potentials and
sensations induced by CO
laser stimulation. Hum Neurobiol 1984;3:33–40.
 Buneo CA, Andersen RA. The posterior parietal cortex: sensorimotor interface
for the planning and online control of visually guided movements.
 Burgess PR, Perl ER. Myelinated afferent ﬁbres responding speciﬁcally to
noxious stimulation of the skin. J Physiol 1967;190:541–62.
 Chudler EH, Dong WK. The assessment of pain by cerebral evoked potentials.
 Colby CL, Goldberg ME. Space and attention in parietal cortex. Annu Rev
 Darwin C. The expression of the emotions in man and animals. London: Murray;
 de Lafuente V, Romo R. Neuronal correlates of subjective sensory experience.
Nat Neurosci 2005;8:1698–703.
 Eidelberg D, Galaburda AM. Inferior parietal lobule. Divergent architectonic
asymmetries in the human brain. Arch Neurol 1984;41:843–52.
 Eimer M, Forster B, Van Velzen J. Anterior and posterior attentional control
systems use different spatial reference frames: ERP evidence from covert
tactile-spatial orienting. Psychophysiology 2003;40:924–33.
 Gallace A, Soto-Faraco S, Dalton P, Kreukniet B, Spence C. Response
requirements modulate tactile spatial congruency effects. Exp Brain Res
 Gallace A, Spence C. The cognitive and neural correlates of ‘‘tactile
consciousness’’: a multisensory perspective. Conscious Cogn 2008;17:
 Gallace A, Spence C. The role of somatosensory cortex in the awareness of
tactile information. Psyche (Stuttg) 2010;16:30–67.
 Grefkes C, Fink GR. The functional organization of the intraparietal sulcus in
humans and monkeys. J Anat 2005;207:3–17.
 Holmes NP, Spence C. Multisensory integration: space, time and
superadditivity. Curr Biol 2005;15:R762–4.
 Iannetti GD, Hughes NP, Lee MC, Mouraux A. Determinants of laser-evoked
EEG responses: pain perception or stimulus saliency? J Neurophysiol
 Jeannerod M, Michel F, Prablanc C. The control of hand movements in a case of
hemianaesthesia following a parietal lesion. Brain 1984;107:899–920.
 Jung TP, Makeig S, Humphries C, Lee TW, McKeown MJ, Iragui V, Sejnowski TJ.
Removing electroencephalographic artifacts by blind source separation.
 Kitazawa S. Where conscious sensation takes place. Conscious Cogn
 Knecht S, Kunesch E, Schnitzler A. Parallel and serial processing of haptic
information in man: effects of parietal lesions on sensorimotor hand function.
 Kupers R, Faymonville ME, Laureys S. The cognitive modulation of pain:
hypnosis- and placebo-induced analgesia. Prog Brain Res 2005;150:251–69.
 Lee MC, Mouraux A, Iannetti GD. Characterizing the cortical activity through
which pain emerges from nociception. J Neurosci 2009;29:7909–16.
 Legrain V, Guerit JM, Bruyer R, Plaghki L. Attentional modulation of the
nociceptive processing into the human brain: selective spatial attention,
probability of stimulus occurrence, and target detection effects on laser
evoked potentials. Pain 2002;99:21–39.
 Legrain V, Perchet C, Garcia-Larrea L. Involuntary orienting of attention to
nociceptive events: neural and behavioral signatures. J Neurophysiol
 Libet B, Alberts WW, Wright Jr EW, Feinstein B. Responses of human
somatosensory cortex to stimuli below threshold for conscious sensation.
 Lloyd DM, Merat N, McGlone F, Spence C. Crossmodal links between audition
and touch in covert endogenous spatial attention. Percept Psychophys
 Macaluso E, Driver J. Multisensory spatial interactions: a window onto
functional integration in the human brain. Trends Neurosci 2005;28:264–71.
 Martin JH. Anatomical substrates for somatic sensation. In: Kandel ER, Schwartz
JH, editors. Principals of neural science. Philadelphia: Elsevier; 1985. p. 301–15.
 Milner AD, Goodale MA. The visual brain in action. Oxford, UK: Oxford
University Press; 1995.
 Miron D, Duncan GH, Bushnell MC. Effects of attention on the intensity and
unpleasantness of thermal pain. Pain 1989;39:345–52.
 Moseley GL, Gallace A, Spence C. Is mirror therapy all it is cracked up to be?
Current evidence and future directions. Pain 2008;138:7–10.
 Moseley GL, Parsons TJ, Spence C. Visual distortion of a limb modulates the
pain and swelling evoked by movement. Curr Biol 2008;18:R1047–8.
 Mountcastle V. The sensory hand: neural mechanisms of somatic
sensation. Cambridge, MA: Harvard University Press; 2005.
 Mouraux A, Iannetti GD. Across-trial averaging of event-related EEG responses
and beyond. Magn Reson Imaging 2008;26:1041–54.
 Mouraux A, Iannetti GD. Nociceptive laser-evoked brain potentials do not
reﬂect nociceptive-speciﬁc neural activity. J Neurophysiol 2009;101:3258–69.
 Pagel B, Heed T, Roder B. Change of reference frame for tactile localization
during child development. Dev Sci 2009;12:929–37.
 Paillard J. Body schema and body image–a double dissociation in deafferented
patients. In: Gantchev GN, Mori S, Massion J, editors. Motor control. Today and
tomorrow. Soﬁa, Bulgaria: Academic Publishing House; 1999. p. 197–214.
 Palva S, Linkenkaer-Hansen K, Naatanen R, Palva JM. Early neural correlates of
conscious somatosensory perception. J Neurosci 2005;25:5248–58.
 Perenin MT, Jeannerod M. Visual function within the hemianopic ﬁeld
following early cerebral hemidecortication in man—I. Spatial localization.
 Polanyi M. Personal knowledge. London: Routledge & Kegan Paul; 1962.
 Preissl H, Flor H, Lutzenberger W, Duffner F, Freudenstein D, Grote E,
Birbaumer N. Early activation of the primary somatosensory cortex without
conscious awareness of somatosensory stimuli in tumor patients. Neurosci
 Ramachandran VS, Rogers-Ramachandran D, Cobb S. Touching the phantom
limb. Nature 1995;377:489–90.
 Schubert R, Blankenburg F, Lemm S, Villringer A, Curio G. Now you feel it—now
you don’t: ERP correlates of somatosensory awareness. Psychophysiology
 Treede RD, Kief S, Holzer T, Bromm B. Late somatosensory evoked cerebral
potentials in response to cutaneous heat stimuli. Electroencephalogr Clin
 Vogel H, Port JD, Lenz FA, Solaiyappan M, Krauss G, Treede RD. Dipole source
analysis of laser-evoked subdural potentials recorded from parasylvian cortex
in humans. J Neurophysiol 2003;89:3051–60.
 Volpe BT, LeDoux JE, Gazzaniga MS. Spatially oriented movements in the
absence of proprioception. Neurology 1979;29:1309–13.
 Weiskrantz L, Warrington EK, Sanders MD, Marshall J. Visual capacity in the
hemianopic ﬁeld following a restricted occipital ablation. Brain 1974;97:
 Yamamoto S, Kitazawa S. Reversal of subjective temporal order due to arm
crossing. Nat Neurosci 2001;4:759–65.
 Ziegler M, Moore E, Wilson M. Comparative accuracy in the localization of
cutaneous pressure and pain. Am J Psychol 1934;46:34–46.
A. Gallace et al. / PAIN
152 (2011) 1418–1423 1423