c o r t e x 4 6 ( 2 0 1 0 ) 7 6 9 – 7 8 0

ava i lab le at www.sc ienced i rec t . com

journa l homepage : www. e lsev ier . com/ loca te / cor tex

Research report

Dissociating cognitive from affective theory of mind:A TMS study

Elke Kalbea,b,*, Marius Schlegelb, Alexander T. Sackc, Dennis A. Nowaka,Manuel Dafotakisa,b, Christopher Bangardd, Matthias Brande,f,Simone Shamay-Tsooryg, Oezguer A. Onura,b and Josef Kesslerb

aInstitute of Neuroscience and Medicine (INM-3), Cognitive Neurology Section, Research Centre Juelich, GermanybDepartment of Neurology, University Hospital of Cologne, GermanycDepartment of Cognitive Neuroscience, Faculty of Psychology, Maastricht University, The NetherlandsdDepartment of Radiology, University Hospital of Cologne, GermanyeDepartment of General Psychology, Cognition, University of Duisburg-Essen, GermanyfErwin L. Hahn Institute for Magnetic Resonance Imaging, Essen, GermanygDepartment of Psychology and Brain and Behavior Center, University of Haifa, Israel

a r t i c l e i n f o

Article history:

Received 5 September 2008

Reviewed 9 December 2008

Revised 6 April 2009

Accepted 9 July 2009

Action editor Elena Rusconi

Published online 29 July 2009

Keywords:

Theory of Mind

Transcranial magnetic stimulation

Dorsolateral prefrontal cortex

5 cm rule

* Corresponding author. Institute of NeuroLeo-Brandt-Str. 5, D-52425 Juelich, Germany

E-mail address: e.kalbe@fz-juelich.de (E.0010-9452/$ – see front matter ª 2009 Elsevidoi:10.1016/j.cortex.2009.07.010

a b s t r a c t

Introduction: ‘‘Theory of Mind’’ (ToM), i.e., the ability to infer other persons’ mental states, is

a key function of social cognition. It is increasingly recognized to form a multidimensional

construct. One differentiation that has been proposed is that between cognitive and

affective ToM, whose neural correlates remain to be identified. We aimed to ascertain the

possible role of the right dorsolateral prefrontal cortex (DLPFC) for cognitive ToM as

opposed to affective ToM processes.

Methods: 1 Hz repetitive transcranial magnetic stimulation (rTMS) was used to interfere

offline with cortical function of the right DLPFC in healthy male subjects who subsequently

had to perform a computerized task assessing cognitive and affective ToM.

Results: RTMS over the right DLPFC induced a selective effect on cognitive but not affective

ToM. More specifically, a significant acceleration of reaction times in cognitive ToM

compared to affective ToM and control items was observed in the experimental (right

DLPFC) compared to the control (vertex) rTMS stimulation condition.

Conclusions: Our findings provide evidence for the functional independence of cognitive

from affective ToM. Furthermore, they point to an important role of the right DLPFC within

neural networks mediating cognitive ToM. Possible underlying mechanisms of the accel-

eration of cognitive ToM processing under rTMS are discussed.

ª 2009 Elsevier Srl. All rights reserved.

science and Medicine (INM-3), Cognitive Neurology Section, Research Center Julich,.Kalbe).er Srl. All rights reserved.

c o r t e x 4 6 ( 2 0 1 0 ) 7 6 9 – 7 8 0770

1. Introduction

Theory of mind (ToM) is defined as the ability to attribute

mental states, such as desires, intentions and beliefs, to other

people in order to explain and predict their behavior (Frith and

Frith, 1999). It constitutes a central aspect of social cognition

which is regarded to be a highly specialized, human-specific

skill that forms a crucial prerequisite to function in social

groups (Adolphs, 2003a, 2003c; Herrmann et al., 2007). ToM is

commonly regarded to be mediated by a complex neural

network including the medial prefrontal cortex (mPFC), the

superior temporal sulcus region, the temporal pole (Frith and

Frith, 2003; Siegal and Varley, 2002), and the amygdalae

(Adolphs, 2003b). Many lesion studies (e.g., Eslinger et al.,

2007; Griffin et al., 2006; Happe et al., 1999; Siegal et al., 1996;

Stuss et al., 2001; Winner et al., 1998) and functional imaging

studies (e.g., Brunet et al., 2000; Gallagher et al., 2000; Sommer

et al., 2007; Vogeley et al., 2001) suggest that ToM and other

social cognitive functions are mediated predominantly by

a network lateralized to the right hemisphere, although

evidence for bilateral (e.g., Vollm et al., 2006; Hynes et al., 2006)

and left-sided involvement also exists (e.g., Baron-Cohen

et al., 1999; Calarge et al., 2003; Channon and Crawford, 2000;

Fletcher et al., 1995; Goel et al., 1995), probably depending on

task type and modality (Kobayashi et al., 2007).

Recent social cognitive neuroscience has begun to define

subcomponents of the complex concept we refer to as ToM.

One important differentiation is that of ‘affective’ versus

‘cognitive’ ToM, although different terms have been used for

these and related concepts (overview in Baron-Cohen and

Wheelwright, 2004; Kalbe et al., 2007). Whereas cognitive ToM,

for example assessed with so-called false belief tasks, is

thought to require cognitive understanding of the difference

between the speaker’s knowledge and that of the listener

(knowledge about beliefs), affective ToM, for example tested

with faux pas and irony tasks, is supposed to require in

addition an empathic appreciation of the listener’s emotional

state (knowledge about emotions) (Shamay-Tsoory et al.,

2006). Brothers (1995, 1997) had postulated a unitary social

‘editor’ which is specialized for processing others’

social intentions but which could not be dissociated into ‘hot’

social cognition (i.e., processing others’ emotional expres-

sions) and ‘cold’ social cognition (i.e., attributing and pro-

cessing cognitive mental states such as beliefs). However,

Eslinger et al. (1996) reported a dissociation between affective

and cognitive aspects of ‘empathy’ in brain damaged patients.

Furthermore, Blair (2005) and Blair and Cipolotti (2000) argued

that divergent results concerning ToM dysfunctions in socio-

pathy may be attributed to a selective deterioration of affec-

tive social cognition (‘emotional empathy’), while individuals

with autism show more difficulties with cognitive than with

emotional empathy. Recently, Shamay-Tsoory and colleagues

found selective deficits of affective as opposed to cognitive

ToM in various patients groups (Shamay-Tsoory and Aharon-

Peretz, 2007; Shamay-Tsoory et al., 2006, 2005).

Already Eslinger (1998) suggested that different regions in

the prefrontal cortex may be relevant for these distinct func-

tions, with a dorsolateral prefrontal cortex (DLPFC) system

mediating cognitive empathy and the orbitofrontal cortex

mediating affective empathy. Shamay-Tsoory et al. (2005)

confirmed the special role of the ventromedial prefrontal

cortex (VMPFC) in processing affective ToM and argued that

cognitive ToM may rather involve both the VMPFC and dorsal

parts of the prefrontal cortex (Shamay-Tsoory and Aharon-

Peretz, 2007). Further confirmation for partially differential

mechanisms in processing affective and cognitive ToM was

recently provided by functional magnetic resonance imaging

(fMRI) studies (Hynes et al., 2006). These studies underline the

particular role of medial and orbital PFC for affective

perspective taking and show involvement of dorsolateral

prefrontal structures for cognitive ToM. Kobayashi et al. (2007)

and Sommer et al. (2007) found involvement especially of the

right-hemispheric DLPFC in false belief tasks (which can be

categorized as cognitive ToM tasks).

In summary, research so far (a) suggests a distinction

between affective and cognitive ToM functions and (b) point to

at least partly different neural correlates mediating these two

subcomponents. However, while the role of the VMPFC for

affective ToM is well documented, neural substrates of

cognitive ToM are less well defined but may include the

DLPFC.

On the basis of the aforementioned considerations, we

aimed to further examine the dissociation of cognitive and

affective ToM processes. We tried to elucidate neural

correlates of cognitive as opposed to affective ToM and,

more specifically, to investigate the functional relevance of

the DLPFC for cognitive ToM performance. For this purpose,

we applied 1-Hz repetitive transcranial magnetic stimula-

tion (rTMS) to the DLPFC of 28 male right-handed healthy

subjects prior to the performance of a computer-based ToM

task that has previously been used to differentially assess

cognitive versus affective ToM (Shamay-Tsoory and

Aharon-Peretz, 2007). Although functional imaging studies

have shown somewhat contradictory results regarding lat-

erality of ToM functions (see above) we decided to perform

rTMS over the right DLPFC for the following reasons: (i) We

used the ‘‘Yoni’’ paradigm introduced by Shamay-Tsoory

and Aharon-Peretz (2007) in which ToM has to be inferred

on the basis of eye gaze and facial expression. According to

Sabbagh (2004), a right-hemispheric mechanism mediates

the decoding of mental states based on immediate infor-

mation, such as eye expression, while a left-hemispheric

network is responsible for complex reasoning about mental

states. It can be speculated that the right-hemispheric

decoding system is utilized when performing the Yoni task

(Shamay-Tsoory and Aharon-Peretz, 2007). (ii) Executive

functions have been conceptualized as a ‘‘co-opted’’ system

for ToM processing (Siegal and Varley, 2002), and recent

functional imaging research points to the central role of the

right DLPFC in executive working memory operations and

cognitive control functions (Lie et al., 2006).

TMS is a well-established tool for inducing transient

changes in brain activity non-invasively in conscious human

volunteers. Over the past couple of years, this ability of

actively interfering with neural processing during behavioral

performance has been increasingly used for the investigation

of causal brain-behavior relations in higher cognitive func-

tions (Pascual-Leone et al., 2000; Sack and Linden, 2003). RTMS

has been applied to different areas within prefrontal cortex in

c o r t e x 4 6 ( 2 0 1 0 ) 7 6 9 – 7 8 0 771

order to successfully interfere with higher cognitive functions

such as visual (Mottaghy et al., 2002; Oliveri et al., 2001) and

spatial (Koch et al., 2005) working memory, verbal and

nonverbal memory encoding (Floel et al., 2004), divided

attention (Wagner et al., 2006), decision making (van’t Wout

et al., 2005), or the implementation of fairness-related

behavior (Knoch et al., 2006a, 2006b). RTMS has been used in

few studies to examine the sensorimotor side of empathy for

pain (Avenanti et al., 2005, 2009). Only one rTMS study

specifically addressed neural correlates of ToM using rTMS,

finding both dorsolateral and temporo-parietal involvement

(Costa et al., 2008). However, no differentiation was made

between cognitive and affective ToM.

For our study, we hypothesized dissociable effects of rTMS

over the right DLPFC on ToM. More specifically, on the basis of

the assumption that the DLPFC is involved in the neural

network which mediates cognitive but not affective ToM, we

expected a selective effect of rTMS over the right DLPFC on

cognitive but not affective ToM processes.

2. Methods

2.1. Sample

Twenty-eight male, right-handed subjects (mean age: 24.0,

standard deviation – SD: 2.7) without neurological or psychi-

atric history were included in the study. All subjects had

completed German high school with the highest degree (Abi-

tur) and currently underwent higher university education in

various fields but not psychology. The study protocol was

approved by the local Ethics committee. All subjects signed

informed consent and underwent a medical safety screening

according to international safety guidelines for the use of TMS

(Wassermann, 1998). Cognitive dysfunction was excluded

with the cognitive screening instrument DemTect (Kalbe et al.,

2004; Kessler et al., 2000), subtest 4 (reasoning) of the German

intelligence test battery ‘‘Leistungsprufsystem’’ (LPS 4, Horn,

1983), and the Trail Making Test A and B (TMT, Reitan, 1979;

Tombaugh, 2004). Mean group scores were 17.4 (SD: 1.1) out of

18 points in the DemTect, C-scores of 7.3 (SD: 1.5) for the LPS 4,

and percentiles of 4.4 (SD: 2.8) and 4.7 (SD: 2.9) for TMT

subtests A and B, respectively.

2.2. ToM tasks

A German version of the ‘‘reading the mind in the eyes’’ test

(Baron-Cohen et al., 2001) was used as a general measure of

ToM abilities. To measure cognitive and affective ToM in the

TMS experiment we used a German modified version of the

‘‘Yoni’’ task introduced by Shamay-Tsoory et al. (2006). It is

based on a task previously described by Baron-Cohen and

Goodhart (1994) and involves the ability to judge mental states

via analysis of verbal cues, eye gaze, and facial expression. In

each of the 60 items presented on a computer screen, a face

named Yoni is shown in the middle with four coloured

pictures in the corners showing either faces or examples of

a semantic category (e.g., animals, fruits). An incomplete

sentence about what image Yoni is referring to is also pre-

sented, and the subject has to judge which of the four stimuli

in the corners best fills the gap of the sentence. The items can

be subdivided into three types of categories with 20 items

each, that is (i) cognitive ToM (cog), (ii) affective ToM (aff), and

(iii) control physical condition (phy), with ten first order and

ten second order items in each category (Fig. 1). While answers

in the physical condition only require analysis of physical

attributes of the character, choices in the cognitive and

affective ToM items require mental inferences based on verbal

cues (contained in the sentences), eye gaze and/or facial

expression. More specifically, in the first order ToM stimuli

Yoni’s mental state about one of the four images in the

corners has to be inferred: Yoni is thinking of . (cog1, German:

Yoni denkt an.), or Yoni loves . (aff1, German: Yoni mag.),

while in the more complex second order ToM items the four

stimuli in the corners consist of faces, and an inference

regarding the interaction between Yoni’s and the other stim-

uli’s mental state is necessary. In the second order cognitive

items with the sentence Yoni is thinking of the . that . wants

(cog2, German: Yoni denkt an die., die . will ), both the verbal

and facial cues are neutral. In the second order affective items

with the sentence Yoni loves the . that . loves, (German: Yoni

mag die., die . mag) both cues are affective. The item sets of

all item subcategories are comparable with regard to sentence

complexity and visual complexity.

The task was programmed with the software PRESENTA-

TION. The total task duration was 10 min and 30 sec. All items

were presented in randomized order for a maximum of 10 sec

during which the subjects had to answer by tapping a button

on the square number keyboard on the right side of the

console. The position of the answer buttons (1, 7, 9, 3) corre-

sponded to the positions of the four stimuli in the corners of

the screen. As soon as subjects answered, a plain white screen

was shown until the end of the 10 sec time interval. Between

these fixed time intervals a black fixation cross on a white

screen was presented for .5 sec. In order to ensure compara-

bility of reaction times (RTs), subjects always had to use the

same finger (right middle or index finger) to respond and

return to the starting position on button 5 in the middle of the

number keyboard after each item. For all items, RTs and

accuracy were registered.

Before rTMS stimulation and administration of the real

test, all subjects received an introduction to the Yoni task with

four explaining slides, and a training that resembled the test

but with only 21 items (7 cognitive, 7 affective, and 7 physical)

not included in the test.

2.3. Magnetic resonance imaging (MRI) localisation ofrTMS target site

Each participant underwent a high resolution whole brain

anatomical MRI scan performed on a whole body 1.5 T

scanner (Achieva 1.5, Philips Medicine Systems, Best, the

Netherlands). This allowed for defining the rTMS target site

based on individual anatomical brain structure. To allow exact

positioning of rTMS over the DLPFC, nifedipine capsules were

sticky-taped over two frontal areas navigated prior to MRI

scanning by two common landmark procedures for the

DLPFC. The first of these two procedures determines DLPFC by

detecting the ‘‘motor hot spot’’ for the abductor pollicis brevis

muscle within the hand area of the primary motor cortex by

Fig. 1 – Item examples of the Yoni ToM task modified from Shamay-Tsoory et al. (2007) used in our TMS experiment.

c o r t e x 4 6 ( 2 0 1 0 ) 7 6 9 – 7 8 0772

single pulse TMS and then moving 5 cm anterior and in

parallel to the midsagital line (George et al., 1995). The second

approach uses the international 10–20 system to localize

DLPFC as corresponding to F4 (Herwig et al., 2003) (see Fig. 2).

The exact individual position of the DLPFC was determined at

the junction of BA 8 and BA 9 caudal to the medial section of

the medial frontal gyrus based on the anatomical brain scan of

Fig. 2 – a. Montreal Neurological Institute (MNI) headmesh show

coordinates. Capsule 1 indicates the stimulation site as determ

Capsule 2 indicates F4, the stimulation site as determined by th

Anatomical regions shown on segmentations of the MNI templ

each participant. This prefrontal section was used because the

dorsal part of the lateral prefrontal cortex is most clearly

related with complex executive functions (Lie et al., 2006;

Miller and Cohen, 2001; Petrides, 2005). Furthermore, this area

has been found to be active during false belief reasoning

which can be conceptualized as a cognitive ToM task (Sommer

et al., 2007).

ing the average locations of the two capsules in Talairach

ined by the 5 cm rule (x [ 51 ± 6, y [ 34 ± 11, z [ 53 ± 7).

e 10–20 system (x [ 46 ± 4, y [ 49 ± 5, z [ 45 ± 6). b.

ate.

c o r t e x 4 6 ( 2 0 1 0 ) 7 6 9 – 7 8 0 773

In order to navigate the rTMS coil to the exact scalp posi-

tion for stimulation of the DLPFC, the location of the DLPFC

was calculated in relation to the anatomical locations

proposed by each landmark procedure in three-dimensional

MRI reconstruction. The final actual rTMS could either be

based on one of the locations indicated by the two landmark

procedures or on a different location on the scalp when both

methods failed to overlie the intended cortical target site. The

advantage of this approach is two-fold: first it provides

a precise and individual determination of the MRI-guided

rTMS target site and second it offers an empirical assessment

of the accuracy and validity of the two most commonly used

standard anatomical landmark approaches for localizing BA 9.

2.4. TMS protocol

A Magstim Rapid2 stimulator (Magstim company, Whitland,

UK), set at 100% of the individual resting motor threshold, and

a 70 mm figure of eight coil were used to deliver a 15 min

single train of 900 1 Hz rTMS at 100% of the motor threshold.

Stimulation parameters were chosen according to the 1 Hz

procedure described by Maeda et al. (2000) which has shown

to result in a 10–15 min reduction of cortical excitability of the

target area. For the detection of the resting motor threshold

the coil was placed tangentially over the right primary motor

cortex at the optimal site for the response of the left first

dorsal interosseus muscle. The resting motor threshold was

defined as the stimulator output intensity that evoked at least

5 out of 10 motor potentials of a minimum amplitude of 100 mV

from the contralateral first dorsal interosseus muscle (mean

was 58.4%, SD: 4%). Each subject received rTMS at two

different locations – one at the cortical target site of right BA9,

and one vertex (Cz) stimulation as control condition (Best-

mann et al., 2002; Koch et al., 2006; Pascual-Leone et al., 1996).

Cz was localized according to the international 10–20 system

(Jasper, 1958). Concerning coil orientation, the figure eight coil

was held tangentially to subjects’ cortex in the angle of motor

spot localization. This corresponded roughly to an angle of 45�

to midsagital line of the subject’s cortex. Holding the coil was

done manually with both hands during the entire stimulation.

2.5. Procedure

The study was conducted as a within-subject design, where

half of the subjects were stimulated at the target area first, and

the other half was stimulated at the control site first. The

sequence of stimulation was randomly assigned to each

participant. Subsequently to the first stimulation the subject

was tested with the Yoni ToM task. After a 30 min inter-

stimulation break the second stimulation was conducted after

which again the Yoni ToM task was administered. ToM testing

started immediately after stimulation. To ensure that subjects

were familiar with the task so that simple learning effects

during test administration under rTMS could be avoided, all

subjects received an introduction and training of the Yoni task

prior to the first stimulation. Furthermore, to ensure that

subjects did not occupy themselves with the experiment at

hand during the 30 min inter-stimulation break they had to

administer a filler task during that break. For this purpose,

a questionnaire (personality questionnaire NEO-FFI, Borkenau

and Ostendorf, 1993) was chosen which was cognitively not

demanding, did not interfere with the experiment, and had an

administration time of approximately 30 min.

2.6. Statistical analysis

All statistical analyses were carried out using the Statistical

Package for the Social Sciences (SPSS) version 15 for Windows

(Release 15.0.0, Chicago: SPSS Inc.). After checking for statis-

tical normal distribution of the data with the Kolmogorov–

Smirnov-Test, a general linear model repeated measures

analysis on the factors ToM condition (cognitive ToM vs

affective ToM vs control physical items of the Yoni task) and

rTMS stimulation condition (experimental vs control) was

employed. For post-hoc testing paired samples t-tests with

corrected a were used.

3. Results

3.1. General ToM abilities

In the ‘reading the mind in the eyes’ task the group reached

a mean of 25.6 (SD¼ 2.1) points (max. score¼ 36) indicating

age- and gender-adequate ToM abilities according to the

normative data provided by Baron-Cohen et al. (2001).

3.2. TMS adverse events

Side effects that occurred due to rTMS stimulation were mild

headache after stimulation in two subjects, eye or nose

twitching during stimulation in 16 subjects and jaw contrac-

tions during stimulation in one subject. One candidate subject

suffered a syncope during motor spot localization after

application of 15 single pulses at different output intensities

with a maximum of 70%. After an Electroencephalography

(EEG) recording with normal results the subject was excluded

from further participation.

3.3. Experimental ToM task ‘‘Yoni’’: RTs

Mean RTs of the main Yoni ToM task categories for the

experimental and control stimulation conditions are indi-

cated in Table 1. Control physical items were processed

significantly faster than cognitive (t¼ 11.223, df¼ 27, p< .001)

and affective (t¼ 11.92, df¼ 27, p< .001) items in the experi-

mental as well as in the control stimulation condition

(t¼ 9.987, df¼ 27, p< .001 for cognitive and t¼ 8.739, df¼ 27,

p< .001 for affective items). Affective items were processed

significantly faster than cognitive items in the experimental

condition (t¼ 11.920, df¼ 27, p< .001) and in the control

condition (t¼ 3.700, df¼ 27, p< .001).

In a general linear model repeated measure analysis, the

factors stimulation site (two stages: experimental vs control)

and item type (three stages: cognitive vs affective vs control),

and the between-subject factor order of condition (experi-

mental – control vs control – experimental) were used, the

latter of which is important to account for possible order

effects. In this analysis, there was a significant main effect for

stimulation site [Pillai’s Trace¼ .262, F(1,27)¼ 9.230, p¼ .005]

Table 1 – Mean RTs in msec of answers to the categories of the Yoni ToM task in the two rTMS conditions.

Control stimulation Experimental stimulation p-value

Mean RT (ms) (SD) Mean RT (ms) (SD)

Cognitive items (total) 2908 (629) 2625 (587) .001*

cog1 1989 (445) 1849 (431) .021

cog2 3827 (936) 3402 (801) .004*

Affective items (total) 2658 (580) 2565 (586) .199

aff1 2130 (530) 2032 (544) .330

aff2 3187 (699) 3096 687 .167

Physical items (total) 1997 (406) 1881 (415) .042

phy1 1707 (368) 1655 (370) .248

phy2 2287 (486) 2107 (488) .028

*p < .05

c o r t e x 4 6 ( 2 0 1 0 ) 7 6 9 – 7 8 0774

and item category [Pillai’s Trace¼ .853, F(2,26)¼ 72.603,

p< .001] and a significant interaction effect between the

factors item category and stimulation site [Pillai’s

Trace¼ .258, F(2,26)¼ 4.337, p¼ .024]. However, neither the

interaction stimulation site with order of condition nor the

interaction item category with order of condition nor the three

way interaction stimulation site with item category with order

of condition were significant [Pillai’s Trace¼ .128,

F(2,27)¼ 3.802, p¼ .062; Pillai’s Trace¼ .036, F(2,26)¼ .471,

p¼ .630; and Pillai’s Trace¼ .111, F(2,26)¼ 1.558, p¼ .230,

respectively]. Thus when stimulated experimentally

compared to control stimulation, subjects differed signifi-

cantly in their RTs between categories, and order of stimula-

tion did not influence this rTMS effect on ToM performance.

Post-hoc paired samples t-test, with a corrected a of .016

between experimental and control stimulation for the item

categories elicited that only RTs in the cognitive ToM category

differed significantly (t¼�3.618, df¼ 27, p¼ .001) (Fig. 3).

These significant differences corresponded to a fastening

of RTs in cognitive ToM items in the experimental stimulation

condition. The delta between the two conditions ranged from

23 to 287 msec across individuals. When subcategories were

analyzed (cog1, cog2, aff1, aff2, phy1, phy2, Table 1) with

paired samples t-test and a corrected a of .008 only RTs in the

Fig. 3 – Reaction time differences control minus

experimental condition for cognitive and affective ToM

items.

cog2 category differed significantly (t¼�3.171, df¼ 27,

p¼ .004) (Fig. 4). For the cog1 category, p was .021.

To analyse whether RTs were stable over the duration of

the task for cognitive ToM items, paired samples t-tests of the

first versus the second half data were performed for each

condition. No significant differences were observed for cog1,

cog 2, and total cognitive ToM items indicating that there were

no learning effects.

3.4. Experimental ToM task ‘‘Yoni’’: accuracy

There were no incorrect answers from any subject. The mean

number of misses (analyzed for all item categories) was 3.5

(SD¼ 3.3) in the experimental and 2.8 (SD¼ 2.9) in the control

condition. Only four out of 28 subjects (14.3%) had no misses

indicating that there was no ceiling effect in performance and

that task difficulty was adequate. A general linear model

repeated measures procedure for misses in the Yoni ToM task

using the factors also included in the RT analysis (i.e., ToM

condition and rTMS stimulation condition) showed no

significant results, even though there was a trend for inter-

action between the factors stimulation site and item category

[Pillai’s Trace¼ .188, F(2,26)¼ 3.009, p¼ .067]. Remarkably,

within-group comparison of misses in the cog2 items in

Fig. 4 – Reaction time differences control minus

experimental condition for subcategories of cognitive,

affective, and physical items.

c o r t e x 4 6 ( 2 0 1 0 ) 7 6 9 – 7 8 0 775

control versus experimental condition did not show a signifi-

cant difference (Wilcoxon test, Z¼�1.447, p¼ .148), indicating

that there was no specific effect in this item subcategory that

might be related to the results of the RT analysis.

4. Conclusion

The main finding of our study is that rTMS over the right

DLPFC has a selective effect on cognitive but not affective ToM

performance. This result is in concordance with the recently

advanced view that these processes are subcomponents of the

complex concept we refer to as ToM and are at least partially

independent (Blair and Cipolotti, 2000; Eslinger, 1998; Eslinger

et al., 1996). Evidence for a functional dissociability of the

independence of cognitive and affective ToM also comes from

patient studies, which show selective deterioration of affec-

tive ToM in patients with ventromedial damage (Shamay-

Tsoory and Aharon-Peretz, 2007; Shamay-Tsoory et al., 2005),

more pronounced dysfunction in affective than in cognitive

ToM in patients with schizophrenia (Shamay-Tsoory et al.,

2006), and also from psychophysiological findings (using skin

conductance responses) in healthy control subjects (Kalbe

et al., 2007). Furthermore, imaging studies have found

partially different networks mediating cognitive and affective

ToM (Hynes et al., 2006; Vollm et al., 2006). Although a side

result of our study, it should be noted in this context that we

found faster RTs for affective than for cognitive ToM items in

both conditions – a finding that is in concordance with ‘‘Yoni’’

results of Shamay-Tsoory and Aharon-Peretz (2007) and also

with behavioral results from a study that used cognitive and

emotional ToM short stories matched in word length (Hynes

et al., 2006). Albeit speculative at this point, the affective items

might be easier than the cognitive items in the Yoni task since

they involve an additional cue for making the decision: a smile

or a frown. This may enhance ToM processing. Alternatively,

the results could also reflect different mechanisms underlying

cognitive and affective ToM. Referring to the two fundamen-

tally different mechanisms that have been proposed to

explain the process of mentalizing, ‘simulation theory’ posits

that other people’s mental states are represented by repli-

cating or mimicking the mental life of the other person and

thus ‘slipping in the other person’s shoes’, while according to

the ‘theory theory’, others’ mental states are modelled ratio-

nally by a knowledge system that is independent from one’s

own mental states (Gallese and Goldman, 1998). Instead of

favouring one of these mechanisms, it has been hypothesized

that both of them exist and that cognitive ToM may primarily

represent a cognitive process which relies on ‘theories’ of

mind corresponding to the ‘theory theory’ while simulation

may rather be the underlying mechanism for affective ToM

(Adolphs, 2002; Adolphs et al., 2000; Heims et al., 2004; Kalbe

et al., 2007; Mitchell et al., 2005; Shamay-Tsoory and Aharon-

Peretz, 2007). Shamay-Tsoory et al. (2005) suggest that simu-

lation mechanism is essential at the beginning of the persons’

affective ToM process and is further used for making infer-

ences regarding the other persons affective mental states.

Affective ToM processing or ‘empathy’ is regarded to rely on

brain structures that develop early in ontogeny including the

limbic system and might thus be mediated by more automatic

and direct neural circuits as compared to cognitive mentaliz-

ing, that could pose more demands on cognitive resources

(Hynes et al., 2006; Mitchell et al., 2005; Satpute and

Lieberman, 2006; Singer, 2006) – and might thus be faster. In

this context it seems to be relevant to consider the connections

between limbic and prefrontal sections. The amygdala, which

is the key structure in evaluating emotional sensory stimuli

(e.g., Phelps, 2006; Phelps and LeDoux, 2005) is both directly

and indirectly connected with the orbitofrontal/ventromedial

part of the frontal lobe (e.g., Brand and Markowitsch, 2006). In

addition, the amygdala is linked to fast automatic responses

via its connections with hypothalamic nuclei and the brain

stem. Amygdala activation can therefore result in fast auto-

nomic arousal (e.g., measured by skin conductance responses),

which is then perceived by somatosensory cortex. Information

about the emotionality of stimuli can significantly influence

evaluative processes, such as decision making, ToM, and other

complex function (Adolphs, 2001, 2003a, 2003b, 2003c; Bechara

et al., 2003; Brand et al., 2007; Damasio, 1994, 1996). This is

most likely the case due to the aforementioned connections

between amygdala and orbitofrontal cortex which has also

been named ‘‘expanded limbic system’’ (Nauta, 1979). It is

hypothesized that this limbic contribution to higher cognitive

functions, in particular within the field of social cognition and

those tasks that depend upon intuitive processes, is linked to

faster reactions, as the emotional system acts fast, parallel,

associative etc. (c.f.; Kahneman, 2003). This may – at least

partially – explain why we found faster reactions to affective

compared to cognitive ToM items. Taken together our results

corroborate the notion that cognitive and affective ToM are

functionally dissociable processes.

RTMS over the right DLPFC in our study induced an accel-

eration of RTs in cognitive ToM, not a decrease as might have

been expected. Certainty about the reliability of this finding

comes from the facts that (1) training effects can be excluded,

as all subjects received a training before test administration so

that they were customized to the task, and more importantly,

RTs for cognitive ToM items were stable over the duration of

the task (2) training or order effects on specific task trials or

items can be excluded, as the order of the items within the

Yoni task as well as the order of rTMS stimulation condition

were randomized across subjects, and also given the result

that there were no statistical effects for the factor order of

condition in the general linear model repeated measure

analysis (3) there was a statistically significant interaction

effect between the factors item category (cognitive vs affective

vs control items) and stimulation site (experimental vs

control). This latter effect stems from a significant difference

of RTs only in cognitive items between experimental and

control stimulation. One possible explanation for the fact that

processing of cognitive ToM items was faster after rTMS over

the DLPFC is that our control stimulation has led to decreased

RTs, not vice versa. However, this is unlikely, as rTMS stim-

ulation over the vertex has been used as control stimulation in

numerous studies using a wide variety of paradigms, and to

the knowledge of the authors has not been shown to have any

specific effect on visual exploration (e.g., Nyffeler et al., 2008)

or other functions (Wiener et al., 2010; Viggiano et al., 2008).

Furthermore, a decrease of RTs after vertex stimulation would

not explain the differential effect on cognitive ToM as

c o r t e x 4 6 ( 2 0 1 0 ) 7 6 9 – 7 8 0776

compared to the affective ToM and control items. Thus the

interpretation that RTs in response to cognitive ToM items

were fastened after rTMS over the right DLPFC seems valid.

One possible explanation for this result is that our stimulation

protocol could have had a facilitating effect when applied over

the right DLPFC and not an inhibitory one when applied over

the primary motor cortex (Maeda et al., 2000). For example,

Sack and Linden (2003) point out that one particular rTMS

protocol can have either inhibitory or facilitatory effects

depending on the cortical area where it is applied and the

behavioral task to be tested. In addition, stimulation charac-

teristics, such as intensity, distribution, depth of penetration,

and accuracy, depend on factors such as scalp-cortex distance

or extent and conductivity of the stimulated tissue. In support

of these considerations, Drager and co-workers found that

specific language (namely picture-word verification) function

was inhibited when a 1 Hz protocol with 600 pulses was con-

ducted on Wernicke’s area and facilitated when it was used on

Broca’s area (Drager et al., 2004). Despite these constraints,

Machii et al. (2006) in their recent review come to the

conclusion that deducing stimulation parameters which are

valid for motor areas and applying them to the study of

cognitive function is the standard procedure which has shown

to produce coherent results. Thus, although general questions

remain regarding the effect of our specific rTMS protocol, it is

definite that our stimulation protocol interfered with normal

processing of ToM in the DLPFC.

Assuming that our rTMS protocol inhibited excitability of

the right DLPFC, the fastening of RTs during the cognitive ToM

tasks suggest that normal functioning of the right DLPFC is

detrimental for performance in cognitive ToM processing.

Thus inhibition of the right DLPFC must have facilitated other

brain regions relevant for task performance, possibly by the

mechanism of ‘‘transcallosal inhibition’’. It is known that low

frequency rTMS has been shown to reduce transcallosal

inhibition within the motor system and may facilitate corti-

cospinal excitability of the not stimulated motor cortex (Gilio

et al., 2003; Pal et al., 2005). 1 Hz rTMS over the primary motor

cortex facilitates function of the contralateral homologue by

reduction of transcallosal inhibition (Kobayashi et al., 2004;

Takeuchi et al., 2005). Comparable effects have also been

demonstrated for higher cortical functions. For example,

hampering function of the relevant left-hemispheric language

areas, either by stroke or after rTMS, causes enhanced neural

activation of the contralateral homotopic areas (Heiss et al.,

2002; Thiel et al., 2006). Also, the processing of specific

emotions such as anger or anxiety known to be lateralized can

be modulated by rTMS over the right PFC (van Honk et al.,

2002). Finally, low frequency rTMS stimulation of the right

frontal cortex is as effective as high frequency rTMS stimu-

lation of the left frontal cortex in patients with depression

(Isenberg et al., 2005).

In context of the task under discussion inhibition of the

right DLPFC by 1 Hz rTMS may have released left DLPFC from

transcallosal inhibition and resulted in enhanced function

within this area. This would point to a left rather than a right-

hemispheric DLPFC relevance for cognitive ToM. There is

evidence for involvement of the left PFC in ToM processing

(e.g., Baron-Cohen et al., 1999; Calarge et al., 2003; Channon

and Crawford, 2000; Fletcher et al., 1995; Gallagher et al., 2000;

Goel et al., 1995). Sabbagh (2004) suggested two anatomically

and functionally different ToM networks in the human cortex:

a right-hemispheric one, especially in the orbitofrontal and

medial temporal cortex, mediating ‘decoding mental states

from outside cues’, and a left-hemispheric network, especially

in the left medial frontal cortex, mediating ‘reasoning about

those mental states’. Left-sided cortical involvement in ToM

processing also includes lateral prefrontal structures (e.g.,

Baron-Cohen et al., 1999; Channon and Crawford, 2000; Sha-

may-Tsoory and Aharon-Peretz, 2007). In line with these

results, Satpute and Lieberman (2006) recently proposed the

framework of a ‘reflexive’ system for automatic social

perception (which relies on limbic/ventromedial and temporal

structures and is needed to code the trait and evaluative

implications of an observed behavior), as opposed to

a ‘reflective’ system for controlled social perception. The latter

system is supposed to be mediated, among other structures,

partly by the lateral prefrontal cortex, which is known to

mediate reasoning and logic, analogy, mathematical problem-

solving as well as working memory and other executive

functions. Satpute and Lieberman (2006) propose that this

reflective system is involved when ‘symbolic computation’ is

necessary in a ToM task. More precisely, the system could

provide a corrective process of automatically generated

hypothesis about interpretations of behavior (mediated by

other structures), i.e., a ‘selection process’ (see also Leslie

et al., 2004, 2005), and is needed where multiple mental

perspectives have to be considered, self knowledge inhibited,

and beliefs considered in relation to subsequent mental states

(Bull et al., 2007).

In line with the aforementioned arguments one may

speculate that rTMS induced inhibition of right DLPFC func-

tioning may cause stronger involvement of emotional reac-

tions to cognitive tasks compared to intact right DLPFC

functions. The DLPFC is connected with other prefrontal areas

(ventrolateral and orbitofrontal sections) and basal ganglia,

via thalamic nuclei (Alexander and Crutcher, 1990; Alexander

et al., 1990; Barbas, 2000; Brand and Markowitsch, 2008) and

DLPFC functioning can inhibit orbitofrontal and limbic acti-

vation involved in social cognition and emotion processing

(for a discussion of disinhibition and prefrontal cortex see

Zamboni et al., 2008). Accordingly an inhibition of the right

DLPFC may result in a disinhibition of orbitofrontal func-

tioning that then facilitates solving cognitive ToM items in

a more emotional and therefore faster way than usually done,

at least as long as the items are not too complex and do not

necessarily involve an executive component.

Although ToM and executive functions can be deteriorated

independently and thus seem dissociable (e.g., Fine et al.,

2001; Lough et al., 2001; Pickup, 2008; Rowe et al., 2001; Stone

et al., 1998), an association between the two has frequently

been shown (e.g., Channon and Crawford, 2000; Kobayashi

et al., 2007; Perner and Lang, 1999; Perner et al., 2002; Sabbagh

et al., 2006). It appears as if executive functions serve as a ‘co-

opted’ system (next to a ‘core’ ToM system), which is neces-

sary to succeed at least in particular variants of ToM tasks

(Siegal and Varley, 2002). Cognitive ToM tasks which require

attributions about the propositional attitudes such as belief,

knowledge, intentions, are more likely to fall into this category

than affective ToM tasks that are associated with the ability to

c o r t e x 4 6 ( 2 0 1 0 ) 7 6 9 – 7 8 0 777

empathize (Shamay-Tsoory et al., 2002) and may involve

implicit affect sharing (Singer, 2006) through simulation pro-

cessing (Mitchell et al., 2005). We thus conclude that the

DLPFC involvement in our study reflects contributions of

executive functions in solving cognitive ToM items as

assessed in the Yoni task. However, when right DLPFC func-

tioning is reduced (via rTMS), integrity of the left DLPFC seems

to be sufficient to deal with the executive component of the

task. In addition, it might be that – in this case – an additional

contribution of limbic structures (i.e., the right orbitofrontal

section), which results from less inhibition by the right DLPFC,

may facilitate solving the cognitive ToM items.

In summary, our study provides empirical evidence for the

functional independence of cognitive and affective ToM.

Furthermore, it points to an important role of the DLPFC

within neural networks mediating cognitive ToM. However,

the exact role of this region within networks mediating ToM

needs to be specified. Future studies are warranted to assess

functional and effective brain connectivity between left and

right DLPFC during the execution of cognitive versus affective

ToM tasks. More concretely, fMRI connectivity studies (Friston

et al., 2003; Goebel et al., 2003) might reveal the exact neuro-

computational mechanisms within bilateral DLPFC during

cognitive versus affective ToM, on the bases of which opti-

mized rTMS protocols could be applied over left versus right

DLPFC in order to further elicit the relevance of this region for

cognitive ToM processes.

Acknowledgements

We thank Michelle Moerel, Faculty of Psychology, Maastricht

University, for support in graphical image processing, and

Ingo Meister and Mitra Ameli, Department of Neurology,

University of Cologne, for assistance in MRI and rTMS.

Furthermore, the work of the first author was funded in part

by the EC-FP6-project DiMI, LSHBCT-2005-512146.

r e f e r e n c e s

Adolphs R, Damasio H, Tranel D, Cooper G, and Damasio AR. Arole for somatosensory cortices in the visual recognition ofemotion as revealed by three-dimensional lesion mapping.Journal of Neuroscience, 20: 2683–2690, 2000.

Adolphs R. Social cognition and the human brain. In Cacioppo JT,Berntson GG, Adolphs R, Carter CS, Davidson RJ,McClintock MK, McEwen BS, Meaney MJ, Schacter DL,Sternberg EM, Suomi SS, and Taylor SE (Eds), Foundations inSocial Neuroscience. Cambridge: MIT Press, 2002: 313–331.

Adolphs R. Cognitive neuroscience of human social behaviour.Nature Reviews. Neuroscience, 4: 165–178, 2003a.

Adolphs R. Investigating the cognitive neuroscience of socialbehavior. Neuropsychologia, 41: 119–126, 2003b.

Adolphs R. Is the human amygdala specialized for processingsocial information? Annals of the New York Academy of Sciences,985: 326–340, 2003c.

Adolphs R. The neurobiology of social cognition. Current Opinion inNeurobiology, 11: 231–239, 2001.

Alexander GE and Crutcher MD. Functional architecture of basalganglia circuits: Neural substrates of parallel processing.Trends in Neurosciences, 14: 55–59, 1990.

Alexander GE, Crutcher MD, and DeLong MR. Basal ganglia-thalamocortical circuits: Parallel substrates for motor,oculomotor, ‘‘prefrontal’’ and ‘‘limbic’’ functions. Progressin Brain Research, 85: 119–146, 1990.

Avenanti A, Bueti D, Galati G, and Aglioti SM. Transcranialmagnetic stimulation highlights the sensorimotor side ofempathy for pain. Nature Neuroscience, 8: 955–960, 2005.

Avenanti A, Minio-Paluello I, Bufalari I, and Aglioti SM. The painof a model in the personality of an onlooker: Influence ofstate-reactivity and personality traits on embodied empathyfor pain. NeuroImage, 44: 275–283, 2009.

Barbas H. Connections underlying the synthesis of cognition,memory, and emotion in primate prefrontal cortices. BrainResearch Bulletin, 52: 319–330, 2000.

Baron-Cohen S and Goodhart F. The ‘‘seeing leads to knowing’’deficit in autism: The Pratt and Bryant probe. British Journalof Developmental Psychology, 12: 397–402, 1994.

Baron-Cohen S, Ring HA, Wheelwright S, Bullmore ET,Brammer MJ, and Simmons A. Social intelligence in thenormal and autistic brain: An fMRI study. European Journalof Neuroscience, 11: 1891–1898, 1999.

Baron-Cohen S, Wheelwright S, Hill J, Raste Y, and Plumb I. The‘‘Reading the Mind in the Eyes’’ Test revised version: A studywith normal adults, and adults with Asperger syndrome orhigh-functioning autism. Journal of Child Psychology andPsychiatry, 42: 241–251, 2001.

Baron-Cohen S and Wheelwright S. The empathy quotient: Aninvestigation of adults with Asperger syndrome or highfunctioning autism, and normal sex differences. Journal ofAutism and Developmental Disorders, 34: 163–175, 2004.

Bechara A, Damasio H, and Damasio AR. Role of the amygdala indecision-making. Annals of the New York Academy of Sciences,985: 356–369, 2003.

Brand M, Grabenhorst F, Starcke K, Vandekerckhove MMP, andMarkowitsch HJ. Role of the amygdala in decisions underambiguity and decisions under risk: Evidence from patientswith Urbach-Wiethe disease. Neuropsychologia, 45: 1305–1317,2007.

Brand M and Markowitsch HJ. Memory processes and theorbitofrontal cortex. In Zald D and Rauch S (Eds), TheOrbitofrontal Cortex. Oxford: Oxford University Press, 2006:285–306.

Bestmann S, Thilo KV, Sauner D, Siebner HR, and Rothwell JC.Parietal magnetic stimulation delays visuomotor mentalrotation at increased processing demands, 2002.

Blair RJ and Cipolotti L. Impaired social response reversal. A caseof ‘acquired sociopathy’. Brain, 123: 1122–1141, 2000.

Blair RJ. Responding to the emotions of others: Dissociating formsof empathy through the study of typical and psychiatricpopulations. Consciousness and Cognition, 14: 698–718, 2005.

Borkenau P and Ostendorf F. NEO-Funf-Faktoren-Inventar (NEO-FFI)nach Costa und McCrae – Deutsche Fassung (nach Costa PT, andMcCrae RR, 1985). Gottingen: Hogrefe, 1993.

Brand M and Markowitsch HJ. The role of the prefrontal cortex inepisodic memory. In Dere E, Huston JP, and Easton A (Eds),Handbook of Episodic Memory. Amsterdam: Elsevier, 2008:317–341.

Brothers L. Neurophysiology of the perception of intentions byprimates. In Gazziniga MS (Ed), The Cognitive Neurosciences.Cambridge: MIT Press, 1995: 1107–1115.

Brothers L. Friday’s Footprint: How Society Shapes the Human Mind.New York: Oxford University Press, 1997.

Brunet E, Sarfati Y, Hardy-Bayle MC, and Decety JA. PETinvestigation of the attribution of intentions with a nonverbaltask. NeuroImage, 11: 157–166, 2000.

c o r t e x 4 6 ( 2 0 1 0 ) 7 6 9 – 7 8 0778

Bull R, Phillips LH, and Conway CA. The role of control functionsin mentalizing: Dual-task studies of theory of mind andexecutive function. Cognition, 1: 7–15, 2007.

Calarge C, Andreasen NC, and O’Leary DS. Visualizing how onebrain understands another: A PET study of theory of mind.The American Journal of Psychiatry, 160: 1954–1964, 2003.

Channon S and Crawford S. The effects of anterior lesions onperformance on a story comprehension test: Left anteriorimpairment on a theory of mind-type task. Neuropsychologia,38: 1006–1017, 2000.

Costa A, Torriero S, Oliveri M, and Caltagirone C. Prefrontal andtemporo-parietal involvement in taking other�s perspective:TMS evidence. Behavioural Neurology, 19: 71–74, 2008.

Damasio AR. Descartes’ Error: Emotion, Reason and the Human Brain.New York: Grosset/Putnam, 1994.

Damasio AR. The somatic marker hypothesis and the possiblefunctions of prefrontal cortex. Philosophical Transactions of theRoyal Society of London – Series B, 351: 1413–1420, 1996.

Drager B, Breitenstein C, Helmke U, Kamping S, and Knecht S.Specific and nonspecific effects of transcranial magneticstimulation on picture-word verification. The European Journalof Neuroscience, 20: 1681–1687, 2004.

Eslinger PJ, Satish U, and Grattan LM. Alterations in cognitive andaffectively based empathy after cerebral damage. Journal of theInternational Neuropsychological Society, 2: 15, 1996.

Eslinger PJ. Neurological and neuropsychological bases ofempathy. European Neurology, 39: 193–199, 1998.

Eslinger PJ, Moore P, Troiani V, Antani S, Cross K, and Kwok S. Oops!Resolving social dilemmas in frontotemporal dementia. Journalof Neurology, Neurosurgery and Psychiatry, 78: 457–460, 2007.

Fine C, Lumsden J, and Blair RJ. Dissociation between ‘theoryof mind’ and executive functions in a patient with early leftamygdala damage. Brain, 124: 287–298, 2001.

Fletcher PC, Happe F, Frith U, Baker SC, Dolan RJ, andFrackowiak RS. Other minds in the brain: A functional imagingstudy of ‘‘theory of mind’’ in story comprehension. Cognition,57: 109–128, 1995.

Floel A, Poeppel D, Buffalo EA, Braun A, Wu CW, and Seo HJ.Prefrontal cortex asymmetry for memory encoding of wordsand abstract shapes. Cerebral Cortex, 14: 404–409, 2004.

Friston KJ, Harrison L, and Penny W. Dynamic causal modelling.NeuroImage, 19: 1273–1302, 2003.

Frith CD and Frith U. Interacting minds - a biological basis. Science,26: 1692–1695, 1999.

Frith U and Frith CD. Development and neurophysiology ofmentalizing. Philosophical Transactions of the Royal Societyof London. Series B, Biological Sciences, 358: 459–473, 2003.

Gallese V and Goldman A. Mirror neurons and the simulationtheory of mind-reading. Trends in Cognitive Sciences, 12:493–501, 1998.

Gallagher HL, Happe F, Brunswick N, Fletcher PC, Frith U, andFrith CD. Reading the mind in cartoons and stories: An fMRIstudy of ‘theory of mind’ in verbal and nonverbal tasks.Neuropsychologia, 38: 11–21, 2000.

George MS, Wassermann EM, Williams WA, Callahan A,Ketter TA, and Basser P. Daily repetitive transcranial magneticstimulation (rTMS) improves mood in depression. NeuroReport,6: 1853–1856, 1995.

Gilio F, Rizzo V, Siebner HR, and Rothwell JC. Effects on the rightmotor hand-area excitability produced by low-frequencyrTMS over human contralateral homologous cortex. TheJournal of Physiology, 551: 563–573, 2003.

Goebel R, Roebroeck A, Kim DS, and Formisano E. Investigatingdirected cortical interactions in time-resolved fMRI data usingvector autoregressive modeling and Granger causalitymapping. Magnetic Resonance Imaging, 21: 1251–1261, 2003.

Goel V, Grafman J, Sadato N, and Hallett M. Modeling otherminds. Neuroreport, 6: 1741–1746, 1995.

Griffin R, Friedman O, Ween J, Winner E, Happe F, and Brownell H.Theory of mind and the right cerebral hemisphere: Refiningthe scope of impairment. Laterality, 11: 195–225, 2006.

Happe F, Brownell H, and Winner E. Acquired ‘theory of mind’impairments following stroke. Cognition, 70: 211–240, 1999.

Heims HC, Critchley HD, Dolan R, Mathias CJ, and Cipolotti L.Social and motivational functioning is not critically dependenton feedback of autonomic responses: Neuropsychologicalevidence from patients with pure autonomic failure.Neuropsychologia, 42: 1979–1988, 2004.

Heiss WD, Thiel A, Kessler J, and Herholz K. Language activationin ischemic stroke and brain tumor: A PET study. Zentralblattfur Neurochirurgie, 63: 133–140, 2002.

Herrmann E, Call J, Hernandez-Lloreda MV, Hare B, andTomasello M. Humans have evolved specialized skills of socialcognition: The cultural intelligence hypothesis. Science, 317:1360–1366, 2007.

Herwig U, Satrapi P, and Schonfeldt-Lecuona C. Using theinternational 10–20 EEG system for positioning of transcranialmagnetic stimulation. Brain Topography, 16: 95–99, 2003.

Horn W. Leistungsprufsystem. Gottingen: Hogrefe, 1983.Hynes CA, Baird AA, and Grafton ST. Differential role of the

orbital frontal lobe in emotional versus cognitive perspective-taking. Neuropsychologia, 44: 374–383, 2006.

Isenberg K, Downs D, Pierce K, Svarakic D, Garcia K, and Jarvis M.Low frequency rTMS stimulation of the right frontal cortex isas effective as high frequency rTMS stimulation of the leftfrontal cortex for antidepressant-free, treatment-resistantdepressed patients. Annals of Clinical Psychiatry, 17: 153–159,2005.

Jasper HH. The ten-twenty electrode system of the InternationalFederation. Electroencephalography and Clinical Neurophysiology,10: 371–375, 1958.

Kahneman D. A perspective on judgment and choice. AmericanPsychologist, 58: 697–720, 2003.

Kalbe E, Kessler J, Calabrese P, Smith R, Passmore AP, Brand M,et al. DemTect: A new, sensitive cognitive screening test tosupport the diagnosis of mild cognitive impairment and earlydementia. International Journal of Geriatric Psychiatry, 19:136–143, 2004.

Kalbe E, Grabenhorst F, Brand M, Kessler J, Hilker R, andMarkowitsch HJ. Elevated emotional reactivity in affectivebut not cognitive components of theory of mind:A psychophysiological study. Journal of Neuropsychology, 1:27–38, 2007.

Kessler J, Calabrese P, Kalbe E, and Berger F. DemTect: Ein neuesScreening-Verfahren zur Unterstutzung derDemenzdiagnostik. Psycho, 26: 243–347, 2000.

Knoch D, Gianotti LR, Pascual-Leone A, Treyer V, Regard M, andHohmann M. Disruption of right prefrontal cortex bylow-frequency repetitive transcranial magnetic stimulationinduces risk-taking behavior. Journal of Neuroscience, 26:6469–6472, 2006a.

Knoch D, Pascual-Leone A, Meyer K, Treyer V, and Fehr E.Diminishing reciprocal fairness by disrupting the rightprefrontal cortex. Science, 314: 829–832, 2006b.

Kobayashi M, Hutchinson S, Theoret H, Schlaug G, and Pascual-Leone A. Repetitive TMS of the motor cortex improvesipsilateral sequential simple finger movements. Neurology,62: 91–98, 2004.

Kobayashi C, Glover GH, and Temple E. Children’s and adults’neural bases of verbal and nonverbal ‘theory of mind’.Neuropsychologia, 45: 1522–1532, 2007.

Koch G, Oliveri M, Torriero S, Carlesimo GA, Turriziani P, andCaltagirone C. rTMS evidence of different delay anddecision processes in a fronto-parietal neuronal networkactivated during spatial working memory. NeuroImage,24: 34–39, 2005.

c o r t e x 4 6 ( 2 0 1 0 ) 7 6 9 – 7 8 0 779

Koch G, Franca M, Albrecht UV, Caltagirone C, and Rothwell JC.Effects of paired pulse TMS of primary somatosensory cortexon perception of a peripheral electrical stimulus. ExperimentalBrain Research, 172: 416–424, 2006.

Leslie AM, Friedman O, and German TP. Core mechanisms in‘‘theory of mind’’. Trends in Cognitive Sciences, 8: 528–533, 2004.

Leslie AM, German TP, and Polizzi P. Belief-desire reasoning asa process of selection. Cognitive Psychology, 50: 45–85, 2005.

Lie CH, Specht K, Marshall JC, and Fink GR. Using fMRI todecompose the neural processes underlying the WisconsinCard Sorting Test. NeuroImage, 30: 1038–1049, 2006.

Lough S, Gregory C, and Hodges JR. Dissociation of socialcognition and executive function in frontal variantfrontotemporal dementia. Neurocase, 7: 123–130, 2001.

Machii K, Cohen D,Ramos-EstebanezC,and Pascual-LeoneA.Safetyof rTMS to non-motor cortical areas in healthy participants andpatients. Clinical Neurophysiology, 117: 455–471, 2006.

Maeda F, Keenan JP, Tormos JM, Topka H, and Pascual-Leone A.Modulationofcorticospinal excitability by repetitive transcranialmagneticstimulation. Clinical Neurophysiology, 111:800–805,2000.

Miller EK and Cohen JD. An integrative theory of prefrontal cortexfunction. Annual Review of Neuroscience, 24: 167–202, 2001.

Mitchell JP, Banaji MR, and Macrae CN. The link between socialcognition and self-referential thought in the medial prefrontalcortex. Journal of Cognitive Neuroscience, 17: 1306–1315, 2005.

Mottaghy FM, Gangitano M, Sparing R, Krause BJ, and Pascual-Leone A. Segregation of areas related to visual workingmemory in the prefrontal cortex revealed by rTMS. CerebralCortex, 12: 369–375, 2002.

Nauta WJH. Expanding borders of the limbic system concept.In Rasmussen T and Marino R (Eds), Functional Neurosurgery.New York: Raven Press, 1979: 7–23.

Nyffeler T, Cazzoli D, Wurtz P, Luthi M, von Wartburg R, Chaves S,et al. Neglect-like visual exploration behaviour after theta bursttranscranial magnetic stimulation of the right posterior parietalcortex. European Journal of Neuroscience, 27: 1809–1813, 2008.

Oliveri M, Turriziani P, Carlesimo GA, Koch G, Tomaiuolo F, andPanella M. Parieto-frontal interactions in visual-object andvisual-spatial working memory: Evidence from transcranialmagnetic stimulation. Cerebral Cortex, 11: 606–618, 2001.

Pal PK, Hanajima R, Gunraj CA, Li JY, Wagle-Shukla A, andMorgante F. Effect of low-frequency repetitive transcranialmagnetic stimulation on interhemispheric inhibition. Journalof Neurophysiology, 94: 1668–1675, 2005.

Pascual-Leone A, Rubio B, Pallardo F, and Catala MD. Rapid-ratetranscranial magnetic stimulation of left dorsolateralprefrontal cortex in drug-resistant depression. Lancet, 348:233–237, 1996.

Pascual-Leone A, Walsh V, and Rothwell J. Transcranial magneticstimulation in cognitive neuroscience–virtual lesion,chronometry, and functional connectivity. Current Opinionin Neurobiology, 10: 232–237, 2000.

Perner J and Lang B. Development of theory of mind andexecutive control. Trends in Cognitive Sciences, 3: 337–344, 1999.

Perner J, Lang B, and Kloo D. Theory of mind and self-control:More than a common problem of inhibition. Child Development,73: 752–767, 2002.

Petrides M. Lateral prefrontal cortex: Architectonic and functionalorganization. Philosophical Transactions of the Royal Society ofLondon – Series B, 360: 781–795, 2005.

Phelps ME. Emotion and cognition: Insights from studies of thehuman amygdala. Annual Review of Psychology, 57: 27–53, 2006.

Phelps ME and LeDoux JE. Contributions of the amygdala toemotion processing: From animal models to human behavior.Neuron, 48: 175–187, 2005.

Pickup GJ. Relationship between theory of mind and executivefunction in schizophrenia: A systematic review.Psychopathology, 41: 206–213, 2008.

Reitan RM. Manual for Administration of Neuropsychological TestBatteries for Adults and Children. Tucson: ReitanNeuropsychological Laboratory, 1979.

Rowe AD, Bullock PR, Polkey CE, and Morris RG. ‘‘Theory of mind’’impairments and their relationship to executive functioningfollowing frontal lobe excisions. Brain, 124: 600–616, 2001.

Sabbagh MA. Understanding orbitofrontal contributions totheory-of-mind reasoning: Implications for autism. Brainand Cognition, 55: 209–219, 2004.

Sabbagh MA, Xu F, Carlson SM, Moses LJ, and Lee K. Thedevelopment of executive functioning and theory of mind.Psychological Science, 17: 74–81, 2006.

Sack AT and Linden DEJ. Combining transcranial magneticstimulation and functional imaging in cognitive brainresearch: Possibilities and limitations. Brain Research Reviews,21: 41–56, 2003.

Satpute AB and Lieberman MD. Integrating automatic andcontrolled processes into neurocognitive models of socialcognition. Brain Research, 1079: 86–97, 2006.

Shamay-Tsoory SG, Tomer R, Yaniv S, and Aharon-Peretz J.Empathy deficits in Asperger syndrome: A cognitive profile.Neurocase, 8: 245–252, 2002.

Shamay-Tsoory SG, Tomer R, Berger BD, Goldsher D, and Aharon-Peretz J. Impaired ‘‘affective theory of mind’’ is associatedwith right ventromedial prefrontal damage. Cognitive andBehavioral Neurology, 18: 55–67, 2005.

Shamay-Tsoory SG, Shur S, Barcai-Goodman L, Medlovich S,Harari H, and Levkovitz Y. Dissociation of cognitive fromaffective components of theory of mind in schizophrenia.Psychiatry Research, 10: 10–18, 2006.

Shamay-Tsoory SG and Aharon-Peretz J. Dissociable prefrontalnetworks for cognitive and affective theory of mind: A lesionstudy. Neuropsychologia, 45: 3054–3067, 2007.

Siegal M, Carrington J, and Radel M. Theory of mind andpragmatic understanding following right hemisphere damage.Brain and Language, 53: 40–50, 1996.

Siegal M and Varley R. Neuronal systems involved in ‘theory ofmind’. Nature Reviews, 3: 463–471, 2002.

Singer T. The neuronal basis and ontogeny of empathy and mindreading: Review of literature and implications for futureresearch.Neuroscienceand BiobehavioralReviews, 30: 855–863,2006.

Sommer M, Dohnel K, Sodian B, Meinhardt J, Thoermer C, andHajak G. Neural correlates of true and false belief reasoning.NeuroImage, 35: 1378–1384, 2007.

Stone VE, Baron-Cohen S, and Knight RT. Frontal lobecontributions to theory of mind. Journal of CognitiveNeuroscience, 10: 640–656, 1998.

Stuss DT, Gallup GG, and Alexander MP. The frontal lobes arenecessary for ‘theory of mind’. Brain, 124: 279–286, 2001.

Takeuchi N, Chuma T, Matsuo Y, Watanabe I, and Ikoma K.Repetitive transcranial magnetic stimulation of contralesionalprimary motor cortex improves hand function after stroke.Stroke, 36: 2681–2686, 2005.

Thiel A, Schumacher B, Wienhard K, Gairing S, Kracht LW, andWagner R. Direct demonstration of transcallosal disinhibitionin language networks. Journal of Cerebral Blood Flow andMetabolism, 26: 1122–1127, 2006.

Tombaugh TN. Trail making test A and B: Normative datastratified by age and education. Archives of ClinicalNeuropsychology, 19: 203–214, 2004.

van’t Wout M, Kahn RS, Sanfey AG, and Aleman A. Repetitivetranscranial magnetic stimulation over the right dorsolateralprefrontal cortex affects strategic decision-making.Neuroreport, 16: 1849–1852, 2005.

van Honk J, Schutter DJ, d’Alfonso AA, Kessels RP, andde Haan EH. 1 Hz rTMS over the right prefrontal cortexreduces vigilant attention to unmasked but not to maskedfearful faces. Biological Psychiatry, 52: 312–317, 2002.

c o r t e x 4 6 ( 2 0 1 0 ) 7 6 9 – 7 8 0780

Viggiano MP, Giovannelli F, Borgheresi A, Feurra M, Berardi N,Pizzorusso T, et al. Disruption of the prefrontal cortex functionby rTMS produces a category-specific enhancement of thereaction times during visual object identification.Neuropsychologia, 46: 2725–2731, 2008.

Vogeley K, Bussfeld P, Newen A, Herrmann S, Happe F, Falkai P,et al. Mind reading: Neural mechanisms of theory of mind andself-perspective. NeuroImage, 14: 170–181, 2001.

Vollm BA, Taylor AN, Richardson P, Corcoran R, Stirling J, andMcKie S. Neuronal correlates of theory of mind and empathy:A functional magnetic resonance imaging study ina nonverbal task. NeuroImage, 29: 90–98, 2006.

Wagner M, Rihs TA, Mosimann UP, Fisch HU, and Schlaepfer TE.Repetitive transcranial magnetic stimulation of thedorsolateral prefrontal cortex affects divided attentionimmediately after cessation of stimulation. Journal ofPsychiatric Research, 40: 315–321, 2006.

Wassermann EM. Risk and safety of repetitive transcranialmagnetic stimulation: Report and suggested guidelines from theInternational Workshop on the Safety of Repetitive TranscranialMagnetic Stimulation, June 5–7, 1996. Electroencephalography andClinical Neurophysiology, 108: 1–16, 1998.

Wiener M, Hamilton R, Turkeltaub P, Matell MS, and Coslett HB.Fast forward: Supramarginal gyrus stimulation alterstime measurement. Journal of Cognitive Neuroscience, 22: 23–31,2010.

Winner E, Brownell H, Happe F, Blum A, and Pincus D.Distinguishing lies from jokes: Theory of mind deficits anddiscourse interpretation in right hemisphere brain-damagedpatients. Brain and Language, 62: 89–106, 1998.

Zamboni G, Huey ED, Krueger F, Nichelli PF, and Grafman J.Apathy and disinhibition in frontotemporal dementia:Insights into their neural correlates. Neurology, 71: 736–742,2008.