Chapter 10

Emotion, Personality, and the Frontal Lobe

Satoshi Umeda

Abstract Previous neuropsychological and functional neuroimaging studies on

human emotion and personality have revealed the involvement of the frontal lobe.
This chapter presents two original studies, a neuroimaging and a neuropsychological
study, focusing on the neural mechanisms for emotional processing and personality.
The first study concentrates on the neural substrates of advanced emotion. Over
the past decade, many studies have primarily focused on basic emotions. In social

S. Umeda (*)
Department of Psychology, Keio University, Mita 2-15-45,
Minato-ku, Tokyo 108-8345, Japan
e-mail: umeda@flet.keio.ac.jp

S. Watanabe and S. Kuczaj (eds.), Emotions of Animals and Humans: 223

Comparative Perspectives, The Science of the Mind,
DOI 10.1007/978-4-431-54123-3_10, © Springer Japan 2013
224 S. Umeda

situations, however, understanding more advanced complex emotions is also important

for maintaining successful communications with others. To date, most of the neu-
roimaging studies on advanced emotions have focused on the neural substrates of
recognizing those emotions, such as “theory-of-mind” reasoning. To determine the
neural and cognitive mechanisms of advanced emotions, it is essential to focus on the
neural substrates of advanced emotion learning, which remain poorly understood.
The goal of the first study was to address this question using functional magnetic
resonance imaging. The following important findings were obtained in the time-
course data of activated brain areas: (1) the medial prefrontal (BA8) and the anterior
cingulate cortex (BA32) were strongly activated in negative-based emotion learning,
and (2) the right dorsolateral prefrontal cortex (BA9/46) was strongly activated
in positive- and negative-based emotion learning. These results suggest that the
prefrontal area is important for acquiring the relationship between social situations
and complex facial expressions, and that these areas make independent contributions
to learning specific emotions.
The second study focuses on the function of the medial prefrontal cortex. Previous
functional neuroimaging studies on “theory of mind” have demonstrated that the
medial prefrontal cortex is involved when subjects are engaged in various kinds of
mentalizing tasks. Although a large number of neuroimaging studies have been
published, a somewhat small amount of neuropsychological evidence supports
involvement of the medial prefrontal cortex in theory-of-mind reasoning. Findings
are presented from a neuropsychological study for two neurological patients with
damage to the medial prefrontal cortex. The results indicated that neither patient
showed impairment on standard theory-of-mind tests and only mild impairments
were seen on advanced theory-of-mind tests. The most striking finding was that
both patients showed personality changes after surgical operations, leading to char-
acteristics of autism and showing a lack of social interaction in everyday life.
Finally, the possible roles of the medial prefrontal cortex are discussed, with empha-
sis on the importance of using multiple approaches to understand the mechanisms
of theory of mind and medial prefrontal functions.

10.1 Basic and Advanced Emotions

Research on the neural basis of emotion and self has come to be the major topic in
social cognitive neuroscience. Many cognitive neuroscientists in this research area
have focused in particular on the functional neuroanatomy involved in the cortical
midline structures. Numerous different domains in self-processing, including moni-
toring, evaluation, agency, and theory of mind, are all more or less involved in mid-
line structures, including the medial prefrontal cortex and anterior cingulate cortex
(Northoff and Bermpohll 2004).
Over the past several years, a number of neuroimaging studies have tried to
clarify the neural mechanisms of “theory of mind,” and have demonstrated that the
medial prefrontal cortex is activated when subjects are engaged in various kinds of
10 Emotion, Personality, and the Frontal Lobe 225

mind-reading or mentalizing tasks (Frith and Frith 1999). Several recent studies in
developmental cognitive neuroscience have reported some evidence that people
with Asperger syndrome or high-functioning autism show weak activations in the
medial prefrontal cortex in comparison with control subjects (Di Martino et al.
2009; Happé et al. 1996; Nieminen-von Wendt et al. 2003). These findings are con-
sidered fundamental evidence for deficits in theory-of-mind abilities among indi-
viduals with Asperger syndrome or high-functioning autism (Frith 2001).
Concerning types of emotion, no universally accepted typology has been defined,
but one widely used division separates basic emotions from more advanced emo-
tions. The emotions of happiness, surprise, sadness, fear, disgust, and anger have
been identified as the six basic emotions (Ekman 1992). By contrast, complex
advanced emotions such as friendliness, pensiveness, thoughtfulness, and melan-
choly are all required to modulate social communication with others or to state
one’s own inner complex mental states (Shaw et al. 2005). In fact, these kinds of
advanced emotions are highly associated with mind-reading abilities, because
understanding advanced emotions in another is a basis for realizing appropriate
theory-of-mind reasoning or empathy-processing for social communications with
others. The “basic-advanced” distinction of emotion is also supported by several
other findings. For instance, Castelli (2005) reported that people with autism under-
stand the basic emotions but fail to recognize some complex emotions.

10.2 Functional Neuroanatomy of Advanced Emotion

Previous functional neuroimaging studies on human emotion over the past decade
have focused primarily on basic emotions, and several neuroimaging studies of
advanced emotion have been reported to date. For instance, previous studies sug-
gested that increasing regret in a gambling task enhanced activity in some brain
areas including the medial orbitofrontal region (Coricelli et al. 2007). However, the
question of how humans learn advanced emotions remains poorly understood at both
behavioral and neural levels. The goal of the first study was to address the question
using a functional magnetic resonance imaging (MRI) method. Illustrated facial
expressions with unknown novel types of emotions were used to examine the learn-
ing processes of advanced emotions at the behavioral and neural levels.
In this study, subjects comprised 18 healthy adults (9 men, 9 women; mean age,
22.2 years; mean duration of education, 14.6 years). As materials for this experi-
ment, six figures (two positive-based, two negative-based, and two neutral-based
learning conditions) were prepared, consisting of one physical situation with time
series and one complex facial expression (Fig. 10.1).
In the learning (MRI scanning) phase, subjects were instructed to lie on the scanner
bed and asked to learn the relationships between the physical situations displayed
above and the facial expressions displayed below. Each stimulus was displayed for
10 s, followed by the cross-hair presentation. Stimulus-onset asynchrony (SOA)
was 50 s. The subjects were also later required to undertake an acquisition test outside
226 S. Umeda

Positive-based condition

Negative-based condition

Neutral-based condition

Fig. 10.1 Experimental stimuli used in the positive-, negative- and neutral-based learning conditions

the scanner. In the behavioral test phase after the MRI scanning, subjects were asked to
make recognition judgments for figures consisting of one physical situation and one
facial expression projected on the computer screen. The recognition test comprised
36 figures, including 12 correct and 24 incorrect combinations.
For acquisition, a 3-T Allegra system (Siemens, Germany) was used to acquire
high-resolution T1-weighted anatomic images and single-shot gradient-echo echo planar
images (EPIs) with blood oxygen level dependent (BOLD) contrast of 35 axial
slices with the following parameters: cubic resolution, 3.5 mm; repetition time, 2 s;
echo time, 30 ms; and flip angle, 90°. Image preprocessing and data analyses were
performed using Statistical Parametric Mapping software (SPM2; http://www.fil.
ion.ucl.ac.uk/~spm/) on a Matlab platform.
For the behavioral results, the mean proportion of correct responses was 0.86 for
the positive-based condition, 0.81 for the negative-based condition, and 0.83 for the
neutral-based condition. Mean reaction time was 4,394 ms for the positive-based con-
dition, 3,678 ms for the negative-based condition, and 3,991 ms for the neutral-based
condition. No significant differences were apparent among these three conditions.
As imaging results, several substantial findings were shown in the time-course
data for activated brain areas during the three conditions, as follows: (1) the poste-
rior medial prefrontal cortex (BA8) and anterior cingulate cortex (BA32) were
strongly activated in the negative-based learning condition (Fig. 10.2); (2) the right
dorsolateral prefrontal cortex (BA9/46) was strongly activated in both positive- and
negative-based learning conditions (Fig. 10.3); (3) bilateral parahippocampal areas
were strongly activated in the positive-based learning condition (Fig. 10.4).
10 Emotion, Personality, and the Frontal Lobe 227

Fig. 10.2 Time course of signal changes for positive-, negative- and neutral-based learning condi-
tions in the posterior medial prefrontal (BA8) and anterior cingulate cortex (BA32)

0.7
0.6
Signal Change (%)

0.5
0.4 Positive
0.3
Negative
0.2
0.1 Neutral
0
T value
-0.1
-0.2
0 6 12 18 24 30 36 42 48

+30 Peristimulus Time (s)

Right DLPFC (BA 9/46)

Fig. 10.3 Time course of signal changes for positive-, negative- and neutral-based conditions in
the right dorsolateral prefrontal cortex (BA9/46)

The next question is whether any difference in brain activation patterns exists
between learning of simple emotion and complex advanced emotion. To answer
this question, activation patterns were compared between simple and complex
emotion conditions. Simple emotion means stimuli including facial expression
with happy eyebrows and happy mouths, or with fearful eyebrows and fearful mouths.
228 S. Umeda

-4
0.4 0.4
Signal Change (%)

Signal Change (%)

0.3 0.3
0.2 Positive 0.2 Positive
Negative Negative
0.1 Neutral 0.1 Neutral
0 0
-0.1 -0.1
0 6 12 18 24 30 36 42 48 0 6 12 18 24 30 36 42 48
Peristimulus Time (s) Peristimulus Time (s)

Left Parahippocampus Right Parahippocampus

Fig. 10.4 Time course of signal changes for positive-, negative- and neutral-based conditions in
right and left parahippocampal areas

These examples are displayed on the top and middle left panels of Fig. 10.1.
Complex emotion means stimuli including facial expression with happy eyebrows
and fearful mouths, or with fearful eyebrows and happy mouths. These examples
are displayed on the top and middle right panels of Fig. 10.1.
The contrast between brain activations during simple and complex emotion
learning indicated that bilateral superior parietal lobules (BA7), bilateral dorsal pre-
motor areas, and the left ventrolateral prefrontal cortex (VLPFC) were greatly acti-
vated in complex emotion conditions rather than in simple emotion conditions
(Fig. 10.5). Interestingly, the superior parietal lobule and dorsal premotor area are
both highly involved in imitation and have been identified as parts of the mirror
neuron system (Brass and Heyes 2005).
These overall results suggest that the prefrontal and medial temporal areas are
essential for acquiring relationships between situations and facial expressions, and
that those areas make independent contributions to learning specific advanced emo-
tions. The findings of this study show that the medial prefrontal cortex (BA8) and
anterior cingulate cortex (BA32) are strongly activated in negative-based conditions
compared with the other two conditions. This suggests that these areas are critical
for acquiring negative-based advanced emotions, which are supposed to be highly
involved in empathetic processing (Singer et al. 2004, 2006). As mentioned before,
the medial prefrontal cortex has been identified as an area with weak activation
among individuals with Asperger syndrome or high-functioning autism. The present
result that this area is involved mostly in negative-based conditions could represent
further evidence for a lack of theory-of-mind ability in people with Asperger
syndrome or high-functioning autism.
10 Emotion, Personality, and the Frontal Lobe 229

Bilateral Superiror
Parietal Lobules (BA 7)

Left VLPFC (BA 10)

T value

+58

Bilateral Dorsal Premotor Areas (BA 6)

Fig. 10.5 Brain areas showing greater activation during complex emotion learning rather than
simple emotion learning

Another important finding was that activations in the right dorsolateral prefrontal
cortex were greater in positive- and negative-based conditions than in neutral-based
conditions. This finding is consistent with previous results that activation in this
area correlates strongly with autonomic activity and activities in the anterior cingu-
late cortex (Critchley et al. 2001, 2003; Critchley 2004; Fukushima et al. 2010).
The present results suggest that bodily responses under the control of the autonomic
nervous system accompany learning associations between physical situations and
facial expressions with emotional valence.
Moreover, bilateral hippocampal areas showed greater activation in positive-
based learning conditions than in the other two conditions. Several previous studies
have reported that the hippocampus is greatly activated during the processing of
positive emotional stimuli, rather than negative emotional stimuli (Britton et al.
2006; Kuchinke et al. 2005; Prohovnik et al. 2004). Recent morphometric studies
using MRI have reported smaller hippocampus volumes in patients with major
depression (for a review, see Davidson et al. 2002). The present result showing that
the hippocampus is greatly involved in positive-based emotion learning is consistent
with the idea that the hippocampus supports positive-based emotional processing.
The contrast between brain activations during simple and complex emotion
learning shows that parts of the mirror neuron system are greatly involved during
complex emotion learning. Activations in these areas suggest that when people
encounter unknown facial expression with a certain situation, parts of the mirror
neuron system are used to understand what the facial expression means by actual
facial imitations (Leslie et al. 2004). Concerning the greater activation found in the
left VLPFC, previous functional imaging data obtained from a false recognition
230 S. Umeda

task showed that this area is highly involved when people recognize the presented
word as a new word that has not previously been encountered (Umeda et al. 2005).
Activation in this area during complex emotion learning thus suggests that under-
standing of unknown facial expression is based on a saliency detection process for
identification as a novel emotion.
The overall findings from the first study provide some evidence that advanced
emotion learning is realized by a combination of neural activities for learning, for
social and empathetic processing, and for the mirror neuron system.

10.3 Neural Substrates of “Theory of Mind”

A number of neuroimaging studies of “theory of mind” or “mentalizing,” the auto-

matic ability to attribute desires and beliefs to other people, have demonstrated that
the medial prefrontal cortex or anterior cingulate (paracingulate) cortex is involved
in various kinds of tasks requiring mentalizing functions (Brunet et al. 2000; Castelli
et al. 2000; Gallagher and Frith 2003; Gallagher et al. 2000; Vogeley et al. 2001).
This area has been recognized as one of the “social brain” areas, together with areas
like the superior temporal sulcus, amygdala, insula, posterior cingulate (retrosplenial
cortex), and fusiform gyrus (Brüne et al. 2003; Wheatley et al. 2007).
Another neuroscientific approach to theory of mind involves neuropsychological
investigations to examine the performance of brain-damaged patients. The initial study
by Happé et al. (1999) reported that following right-hemisphere damage, patients
showed impaired understanding of materials requiring attribution of mental states.
Several other studies examining the effects of brain lesions on theory-of-mind
performance have reported that focal frontal lesions impair the ability to infer mental
states of others (Bach et al. 2000; Happé et al. 2001; Shamay-Tsoory et al. 2003;
Stuss et al. 2001). More careful examinations to understand the roles of different
areas within the frontal lobe have indicated that right ventromedial prefrontal lesions
impair detection of deception (Stuss et al. 2001) and empathy processing (Shamay-
Tsoory et al. 2003), whereas lesions in the internal capsule impair advanced theory-
of-mind performance for understanding the thoughts and feelings of fictional
characters (Happé et al. 2001). Conversely, other studies have suggested that the ability
to understand mental states was found to be intact in patients with orbitofrontal
lesions (Bach et al. 2000).
Another comprehensive neuropsychological study of over 30 patients with uni-
lateral (right or left) frontal lobe lesions found that both groups exhibited impaired
performance on first- and second-order false belief tests (Rowe et al. 2001). Stone
et al. (1998) tested performance on first- and second-order false-belief tests and the
faux-pas recognition test in patients with orbitofrontal damage, and found that bilateral
orbitofrontal lesions resulted in difficulty only in the ability to recognize a faux pas
(see Sect. 10.6 for details of the tasks). Although these studies focused on the effects
of damage to the frontal lobe, it remains unclear as to whether selective damage to
the medial prefrontal cortex yields any impaired performance on various kinds of
theory-of-mind tests.
10 Emotion, Personality, and the Frontal Lobe 231

Some previous studies have actually examined the performance of patients with
selective damage to the medial prefrontal cortex. Baird et al. (2006) tested two
patients and reported intact intellectual, memory, and language abilities, and visuo-
perceptual functions, but showed weak or impaired performance on selective execu-
tive function tests. No theory-of-mind performance was tested in this study. Another
neuropsychological study addressed the question of theory of mind impairment by
testing a patient with a selective lesion in the medial prefrontal cortex (Bird et al.
2004). They carefully examined performance of the patient on various kinds of
theory-of-mind tests, but found no significant impairment on tests, and thus stated
that extensive medial frontal regions are not necessary for theory-of-mind perfor-
mance. The findings from both studies have some important implications for our
understanding of the effects of damage to the medial prefrontal cortex. Most inter-
esting was the finding that following the damage to this area, patients did not show
any severe impairment on test performance, including theory-of-mind performance.
However, the data are currently too limited to reach solid conclusions on the effects
of medial prefrontal damage.

10.4 Personality Change After Damage to the Frontal Lobe

Another unresolved question is concerned with personality changes in human

patients with damage to their frontal lobes. In the classic case of Phineas Gage,
damage to the orbitofrontal cortex resulted in severe behavioral disturbances in
everyday life (Damasio et al. 1994; Harlow 1848). A number of previous studies
have reported that patients with damage to the ventromedial prefrontal cortex
(VMPFC) show severe sociopathic personality change (Eslinger and Damasio 1985;
Saver and Damasio 1991). Following a lesion of the VMPFC, most patients show
inappropriate social behavior including emotionally insensitive social interaction,
unexpected wandering, alcohol abuse, and confabulation. Despite these sociopathic
psychopathological changes, higher-order cognitive functions including general
memory performance are largely preserved.
Several hypotheses have already been presented to explain the function of
VMPFC. The somatic marker hypothesis proposes that the damage to the VMPFC
precludes the ability to use somatic (emotional) signals that are necessary for guiding
decisions in a positive direction (Bechara et al. 1994, 1996, 1997, 1999; Damasio
1996). This interpretation, based on findings obtained from the gambling task, leads
to the conclusion that the VMPFC mediates decision making. The VMPFC appears
to play an important role in social reasoning (Adolphs et al. 1996). The researchers
using the Wason selection task found that VMPFC patients show impairment in the
reasoning task including familiar cover stories with social laws, whereas they showed
no abnormality when reasoning about more abstract material without a cover story.
This pattern of results differed from that of normal subjects, i.e., the VMPFC patients
had a deficiency in the thematic effect. Adolphs (1999) stated that the VMPFC guides
social reasoning and decision-making by the elicitation of emotional states that serve
to bias cognition. Furthermore, a recent study of functional MRI using a simple
232 S. Umeda

card-playing task indicated that activation of the VMPFC increases as probabilistic

contingencies become more complex (Elliott et al. 1999). Elliott and colleagues
concluded that the VMPFC mediates guessing.
On the other hand, Schacter et al. (1996) reported that one patient (B.G.) exhib-
ited a striking pattern of false recognition after an infarction of the right frontal lobe
including the inferior prefrontal cortex. One possible interpretation for this result is
that B.G.’s recognition decision was based on a fuzzy general event description, or
a “false gist” which was generated from the presented words.
These accounts for the function of the VMPFC appear to be largely compatible,
because a certain type of emotion-based memory processing under complex situa-
tions is more or less required in decision making, social reasoning, and guessing.
It is notable that emotions pertaining to the VMPFC are assumed to be negative
feelings (e.g., fear, anger), because the VMPFC has a direct neural connection to the
amygdala (Tranel 2000). Past studies that focus on the function of the amygdala and
its related circuits indicate that the amygdala will provide a quick and automatic
bias with respect to those aspects of the response that pertain to evaluating the
potentially threatening nature of a situation (Adolphs 1999).
However, the question of whether patients with lesions involving the medial
prefrontal cortex, located in the vicinity of the VMPFC, shows any personality
change remains unanswered. A recent meta-analysis of 39 functional imaging studies
for autism-spectrum disorders indicated that the medial prefrontal cortex was less
activated during social task performance (e.g., theory-of-mind tests) in autism-
spectrum disorders compared with neurotypical controls (Di Martino et al. 2009).
Patients with damage to the medial prefrontal cortex may show personality changes,
leading to characteristics of autism.
Although previous neuropsychological studies of patients with damage to the
medial prefrontal cortex have shed light on the functions of this area, the number of
studies remains limited, and actual effects on social functions following damage to
this area remain poorly understood (Gallagher and Frith 2003). The second study of
this chapter presents two cases of patients with damage to the medial prefrontal
cortex, and reports on their performance on various kinds of theory-of-mind tests,
whether personality changes were evident, and behavioral disturbances in daily
activities.

10.5 Neuropsychological Investigations

The two patients had medial prefrontal damage (Umeda et al. 2010). The first case,
T.O., was a 31-year-old man. A full-time employee of a big electronic company in
Japan, he had undergone neurosurgery for brain tumor. Magnetic resonance imag-
ing revealed that damage extended through the left-dominant medial prefrontal and
anterior cingulate cortices, reaching the left supplementary motor area. The right
hand and leg were moderately paralyzed for a few months after surgery, but these
symptoms later resolved.
10 Emotion, Personality, and the Frontal Lobe 233

The most striking aspect of T.O. was a reported change in personality. According
to his self-report, he noticed that his sense of reality was attenuated after surgery,
leading him to feel detached from the world despite being sure of his location, and
this feeling often occurred in a manner similar to a panic attack. These symptoms
partly resembled the characteristics of depersonalization. He also mentioned that
surgery had made him feel depressed, anxious, and withdrawn from everything.
The second patient, H.C., was a 56-year-old man. An employer of a small private
company in Japan, he had undergone surgery following rupture of a right peri-
callosal artery aneurysm. Magnetic resonance imaging revealed that the area of
damage included the right-dominant medial prefrontal and anterior cingulate cortices,
extending slightly into the right supplementary motor area.
According to self-reports, he noticed that his memory had deteriorated after
surgery, with a feeling that most daily episodes could not be clearly remembered.
He reported difficulty doing two things simultaneously, and became aware that
everything needed a strong effort to be done. He also mentioned that his personality
had changed after surgery, leading him to notice that feelings of sadness and anger
had been dimmed and that he had become much more depressive, anxious, and
withdrawn compared to his previous personality.
Three months after surgery, six of the more frequently used neuropsychological
assessments were conducted to examine higher-order cognitive functions in these
patients. (1) Wechsler Adult Intelligence Scale-Revised (WAIS-R) for general intel-
ligence, (2) Wechsler Memory Scale-Revised (WMS-R) for memory and attention,
(3) Rey Auditory–Verbal Learning Test (RAVLT) for verbal recall ability, (4) Rey–
Osterrieth Complex Figure Test (ROCFT) for visuoconstructive skills and visual
memory, (5) Wisconsin Card Sorting Test (WCST) for abstract reasoning and ability
to appropriately shift cognitive strategies, and (6) Stroop Test for selective attention
and inhibition.
On the intelligence test, T.O., with left-sided damage, showed dissociation
between verbal and performance intelligence quotient (IQ), with an inferior score on
performance IQ. In the IQ test, he showed difficulty in performing subtests of block
design, object assembly, and digit symbols. In terms of memory performance, T.O.
showed lower scores on some measures for identifying delayed recall performance
(e.g., delayed recall on WMS-R, RAVLT, and ROCFT) compared with his relatively
higher scores on other measures. The results of mild amnesia were consistent with
his self-report regarding daily activities. For instance, he reported often becoming
confused in remembering whether he had taken his medication. T.O.’s performances
on tests for executive function were all within the normal range. Performance on the
standardized aphasic test did not show any difficulties in language activities.
The second patient, H.C. with right-sided damage, showed an intellectual perfor-
mance within the normal range. In terms of memory performance, he exhibited nor-
mal scores on WMS-R, although scores on RAVLT and ROCFT were somewhat
lower. In fact, in terms of daily activities he showed difficulty with temporal-order
judgments for everyday episodes within a time range of a few days. H.C.’s perfor-
mance on tests for executive function was also within the normal range. Performance
on the standardized aphasic test did not show any difficulties in language activities.
234 S. Umeda

10.6 Theory-of-Mind Performance After Brain Injury

To clarify theory-of-mind performance, patients T.O. and H.C. were tested using
four types of story comprehension task: (1) first-order false-belief test (Baron-Cohen
et al. 1985; Frith and Frith 1999), (2) second-order false-belief test (Baron-Cohen 1989),
(3) strange stories test (Happé 1994), and (4) faux-pas recognition test (Baron-
Cohen et al. 1999).
The first-order false-belief test is one of the most famous tests for theory-of-mind
reasoning. This test assesses the ability to recognize that others can have false beliefs
about the world that can differ from reality, and that people’s behaviors can be
predicted by the representation of others’ mental states. The more complex second-
order false-belief test requires participants to understand a second person’s concerns
about the world, based on social interactions of minds in which people are con-
cerned about each other’s mental states.
The last two tests were used to examine the more advanced theory-of-mind rea-
soning ability in the patients. The strange stories test assesses the ability to infer
mental states in a story context for social understanding. A previous study reported
that subjects with autism-spectrum disorders show impaired provision of context-
appropriate mental state explanations for strange stories, in comparison with normal
control subjects (Happé 1994).
As well as the strange stories test, the faux-pas recognition test was used to
assess the ability to recognize inappropriate statements in a story context (Baron-
Cohen et al. 1999). Subjects were presented with each story and asked whether a
faux pas was contained. If a faux pas was detected, they were then asked for an
explanation of it. Each subject was also requested to answer two additional ques-
tions to test story comprehension in each story, to see whether a comprehensive
understanding of each story was achieved. Baron-Cohen et al. (1999) reported that
subjects with autism-spectrum disorders show impaired detection of faux pas on the
faux-pas recognition test compared with normal control subjects, despite intact
story comprehension.
In addition, T.O. and H.C. were required to complete all 50 items in the Autism-
Spectrum Quotient (AQ) questionnaire (Baron-Cohen et al. 2001). This question-
naire was developed as a self-administered method of screening for adults with
normal intelligence and traits associated with autism-spectrum disorders. Score
ranges from 0 to 50 in the questionnaire. Adults with Asperger syndrome or high-
functioning autism show a mean score of 35.8, significantly higher than controls
with a mean score of 16.4 (Baron-Cohen et al. 2001). Another recent study has
shown that the threshold score for suspected Asperger syndrome or high-functioning
autism is 26.0 (Woodbury-Smith et al. 2005).
Both patients passed the first- and second-order false-belief tests, providing
expected answers suggesting a proper understanding of each story. In terms of
advanced theory-of-mind tests, both patients showed good performance on the strange
stories test. Overall percentage of providing appropriate explanations for given stories
was 85.7% for T.O. and 100% for H.C. on the strange stories test. On the faux-pas
10 Emotion, Personality, and the Frontal Lobe 235

recognition test, the percentage of detecting faux pas and having appropriate
explanations was 60.0% for T.O. and 100% for H.C. for the provided faux-pas stories.
By contrast, the percentage of detecting “no” faux pas was 70.0% for T.O. and
57.1% for H.C. for the provided control stories. Both patients thus reported faux pas
even in control stories without any faux pas. In addition, both patients showed
higher scores on the two questions for story comprehension (92.5% for T.O. and
89.3% for H.C.).
In terms of the AQ, scores were 31 for T.O. and 29 for H.C., above the threshold
score for Asperger syndrome or high-functioning autism of 26 as defined by
Woodbury-Smith et al. (2005). Interestingly, both patients spontaneously reported
just after completing the questionnaire that they were sure that some of the person-
ality traits focused upon in the questionnaire identified the actual personality changes
they felt. They were then asked to complete the AQ again for what they supposed
their original personality was before surgery. AQ scores for purported presurgical
state of the two patients were 13 for T.O. and 23 for H.C., much lower than the initial
scores and below the threshold score for Asperger syndrome or high-functioning
autism. Both patients thus appear to have developed some autistic personality traits
following surgical operations.
A general finding was that the two patients had developed some characteristics
of autism after surgical operations. To specify these characteristics in greater detail,
these items were compared with items identified in a two-factor structure model
(Hoekstra et al. 2008). Two factors were identified among all 50 items in the AQ,
namely “social interaction” and “attention to detail”. For T.O., all 18 items fell into
the “social interaction” factor while for H.C. six of seven items fell into the “social
interaction” factor. The only item falling into the “attention to detail” factor for H.C.
was the item that showed he had become fascinated by dates.
In another opportunity separate from this study, 11 patients with damage to other
parts of the brain (orbitofrontal lesion, basal forebrain lesion, dorsolateral prefrontal
lesion, medial temporal lesion, amygdala lesion, and traumatic brain injury) were
asked to complete the AQ, to compare scores and possible personality changes
detected by the AQ. Mean score for the 11 patients was 17.0 (range, 9–25), and
median score was 17.0. All patients declared that personality traits identified on the
AQ were unchanged after surgical operations or closed-head injuries.

10.7 Acquired Autism Trait After Medial Prefrontal Damage

The two patients displayed damage basically limited to the medial prefrontal cortex
and showed mild difficulties in memory performance in daily activities, but no serious
problems in language activities and executive functions. These patterns of results
are basically consistent with previous case studies regarding damage to the same
area of the brain (Baird et al. 2006; Bird et al. 2004). Concerning theory-of-mind
tests, performance in the first- and second-order false-belief test was perfect in both
cases. Performance in the advanced theory-of-mind tests by T.O. was slightly
236 S. Umeda

impaired regarding the provision of appropriate explanations in the strange stories

test, and was considerably impaired in the identification of inappropriate verbal
expressions for given contexts in the faux-pas detection test. By contrast, perfor-
mance in advanced theory-of-mind tests by H.C. was not impaired at all in either
test. An interesting finding in both cases was that a faux pas was often reported even
in control stories without any obvious faux pas.
The most notable finding was that both patients showed some difficulties on the
faux-pas recognition test. The percentage of detecting faux pas and providing appro-
priate explanations was 100% for H.C., compared with 60.0% for T.O. Lower per-
formance by T.O. may have been caused by deficits in delayed recall performance
as found in WMS-R and ROCFT. Although T.O. showed higher scores on the two
questions for story comprehension (92.5%), he experienced difficulty recalling the
exact story contents. In fact, T.O. reported the presence of faux pas for all ten faux-
pas stories, but could not recall what the exact contents were in each story. Taking
these facts into account, his basic performance for detecting faux pas may not have
been greatly reduced.
By contrast, both subjects sometimes incorrectly reported faux pas even in stories
containing no faux pas, although they could correctly recognize faux pas in stories
containing faux pas. Various reasons could explain this pattern of results. First, this
pattern could result from perseveration of response in both cases. The response
for detecting faux pas could be a prevailing response, as half of the questions in the
faux-pas test did contain faux pas. However, no strong evidence of perseveration was
found in either case, since very few total perseveration errors were found on WCST.
Second, the pattern could result from general difficulty in understanding global con-
texts in complex situations. If this were the case, the subjects would show some
problems in detecting faux pas in stories containing faux pas. However, the results
were in direct opposition to this prediction. A final possibility is overcompensation.
In a psychiatric sense, this is often defined as an attempt to overcome an actual defect
or unwanted trait by exaggerating in the opposite direction. Self-reports from the two
patients indicated that personality change extended to abnormal feelings in some
emotional dimensions. Unfortunately, no questionnaires examining anxiety traits
were conducted, even though both subjects reported anxiety after surgery. These
changes may have resulted in the subjects being more sensitive to verbal expressions
in comparison to before surgery. The explanation of overcompensation is considered
the most plausible for understanding the overdetection of faux pas.
The most interesting finding was that the two patients showed personality changes
after surgery, resulting in some characteristics of autism. These tendencies were
mainly clarified by findings from the AQ questionnaire. According to the self-
reports, both patients showed a lack of theory-of-mind ability in everyday life,
reduced spontaneous seeking to communicate with others after surgery, and obses-
sive focus on a single subject. To elucidate greater detail of those characteristics,
these items were compared with the items identified in a two-factor structure model
(Hoekstra et al. 2008). As a result, 25 items among the total 26 items for the devel-
opment of autistic personality traits after surgical operations in both patients fell
into the “social interaction” factor. This basically identified acquired functional
10 Emotion, Personality, and the Frontal Lobe 237

deficits following damage to the medial prefrontal cortex as a lack of social

interaction. Surprisingly, H.C. even reported becoming fascinated by dates, which
is considered a strong characteristic of autism. Results for the number of patients
with damage to other areas besides the medial prefrontal cortex revealed that these
personality changes resulted from damage to the medial prefrontal cortex alone.
However, limitations exist to the interpretation of the present results. Patients
were asked to fill out the same questionnaire (AQ) twice, and were requested on the
second trial to answer from the perspective of their previous personality before sur-
gery. This obviously represents a “retrospective report” in the postoperative period,
and the data are clearly of questionable validity. However, the second AQ trial based
on self-reports revealed that some personality traits identified by the questionnaire
matched well with actual personality changes reported after surgery. This suggests
that results of the second trials were substantially valid. The results were consistent
with previous imaging studies for Asperger syndrome, showing that the medial pre-
frontal cortex is highly involved in understanding theory-of-mind stories compared
with understanding control stories in normal control and Asperger syndrome groups,
although the level of peak activation was lower in the Asperger group (Happé et al.
1996). Another interpretation of the AQ rise in both patients is the effect of increased
depression and/or anxiety. T.O. and H.C. both mentioned feeling depressive and
anxious in everyday life. Depression and/or anxiety alone may increase the AQ
score. However, most of the control cases with damage to other areas besides the
medial prefrontal cortex reported feeling more or less depressed and anxious, but
did not show any increase in AQ score after the damage. This evidence suggests
that depression or anxiety alone may not greatly affect AQ score.
Previous studies have reported functional abnormality in the medial prefrontal
cortex in autism-spectrum disorders during social task performance, such as theory-
of-mind reasoning (Di Martino et al. 2009). Based on the model by Hoekstra et al.
(2008), autism personality traits detected in AQ were divided into “social interac-
tion” and “attention to detail” factors. Considering the result that nearly all items for
the development of autistic personality traits in both patients fell into the “social
interaction” factor, the medial prefrontal area does not seem to be involved in the
personality trait for “attention to detail.” Functional abnormality in the medial pre-
frontal cortex in autism-spectrum disorders is likely to be associated with a lack of
social interaction.
The present study remains preliminary, but some essential implications help in
understanding the possible roles of the medial prefrontal cortex. More research is
evidently required to confirm the hypotheses discussed in this study. Besides theory-
of-mind functioning, several recent neuroimaging studies have shown that the
medial prefrontal cortex is involved in moral judgment (Greene et al. 2001), self-
referential processing (Kelley et al. 2002; Schaefer et al. 2006), memory for self
(Macrae et al. 2004), and detecting the communicative intentions of others (Kampe
et al. 2003). If our discussion is expanded by extending regions of interest from the
medial prefrontal cortex to the adjacent anterior cingulate cortex, arguments could
be made from the perspectives of cognitive control (MacDonald et al. 2000), error
detection, or online monitoring (Carter et al. 2000).