Effects of contextualized emotional conflict control on domain-general control: fMRI evidence of neural network reconfiguration

Abstract

Domain-general conflict control refers to the cognitive process in which individuals suppress task-irrelevant information and extract task-relevant information. It supports both effective implementation of cognitive conflict control and emotional conflict control. The present study employed functional magnetic resonance imaging (fMRI) and adopted an emotional valence conflict task and the arrow version of the flanker task to induce contextualized emotional conflicts and cognitive conflicts, respectively. The results from the conjunction analysis showed that the multitasking-related activity in the pre-SMA, bilateral dPMCs, the left pIPS, the left aIPS, and the right inferior occipital gyrus (IOG) represents common subprocesses for emotional and cognitive conflict control, either in parallel or in close succession. These brain regions were used as nodes in the domain-general conflict control network. The results from the analyses on the brain network connectivity patterns revealed that emotional conflict control reconfigures the domain-general conflict control network in a connective way, as evidenced by different communication and stronger connectivity among the domain-general conflict control network. Together, these findings offer the first empirical-based elaboration on the brain network underpinning emotional conflict control and how it reconfigures the domain-general conflict control network in interactive ways.

Full text

Effects of contextualized emotional conict control on
domain-general conict control: fMRI evidence of neural
network reconguration
Tingting Guo,1,2 Xiyuan Wang,1,2 Junjie Wu, 3 W. John Schwieter,4,5 and Huanhuan Liu 1,2
1Research Center of Brain and Cognitive Neuroscience, Liaoning Normal University, Dalian 116029, China
2Key Laboratory of Brain and Cognitive Neuroscience, Dalian, Liaoning Province 116029, China
3Key Research Base of Humanities and Social Sciences of the Ministry of Education, Tianjin Normal University, Tianjin 300382, China
4Language Acquisition, Multilingualism, and Cognition Laboratory/Bilingualism Matters, Wilfrid Laurier University, Waterloo N2L3C5, Canada
5Department of Linguistics and Languages, McMaster University, Hamilton L8S4L8, Canada
Correspondence should be addressed to Huanhuan Liu. Research Center of Brain and Cognitive Neuroscience, Key Laboratory of Brain and Cognitive
Neuroscience, Dalian, Liaoning Province 116029, China. E-mail: abcde69503@126.com.
Abstract
Domain-general conict control refers to the cognitive process in which individuals suppress task-irrelevant information and extract
task-relevant information. It supports both effective implementation of cognitive conict control and emotional conict control. The
present study employed functional magnetic resonance imaging and adopted an emotional valence conict task and the arrow version
of the anker task to induce contextualized emotional conicts and cognitive conicts, respectively. The results from the conjunction
analysis showed that the multitasking-related activity in the pre-supplementary motor area, bilateral dorsal premotor cortices, the
left posterior intraparietal sulcus (IPS), the left anterior IPS and the right inferior occipital gyrus represents common subprocesses for
emotional and cognitive conict control, either in parallel or in close succession. These brain regions were used as nodes in the domain-
general conict control network. The results from the analyses on the brain network connectivity patterns revealed that emotional
conict control recongures the domain-general conict control network in a connective way as evidenced by different communication
and stronger connectivity among the domain-general conict control network. Together, these ndings offer the rst empirical-based
elaboration on the brain network underpinning emotional conict control and how it recongures the domain-general conict control
network in interactive ways.
Keywords: emotional conict control; domain-general conict control; fMRI; euSEM; effective connectivity
© The Author(s) 2024. Published by Oxford University Press.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which
permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
Introduction
Processes of suppressing/controlling conicts, whether cognitive
or emotional, are typical in communication and the ability to
effectively control conict helps humans to better adapt to the
environment around them. The impairment of emotional conict
control particularly has been manifested in a variety of psychi-
atric disorders (Wang et al., 2008; Etkin et al., 2010). Exploring the
neural basis of cognitive conict control and emotional conict
control is of great signicance to explain these psychological pro-
cesses and can have important implications for patients in clinical
settings.
Domain-general conict control refers to a common subpro-
cess which implements efcient performance of cognitive con-
ict control (Ridderinkhof et al., 2004; Gilbert and Burgess, 2008)
and emotional conict control (Etkin et al., 2006). It is a pro-
cess by which individuals extract task-relevant information and
suppress task-irrelevant information when confronted with con-
icting input. There has been a surge of research examining the
neural overlap between emotional conict control and cognitive
conict control (Torres-Quesada et al., 2014; Chen et al., 2018),
and it is reported that different types of conict control recruit
a domain-general conict control network in which each brain
region plays its own role (Ochsner et al., 2009; Xu et al., 2016). For
example, the pre-supplementary motor area (pre-SMA) is a criti-
cal brain region involved in control across emotional and cognitive
conict control (Xu et al., 2016; Xing et al., 2023). This region is also
thought to be associated with conict monitoring, in which sig-
nals are sent to other regions when a conict is identied (Garavan
et al., 2003; Iannaccone et al., 2015).
It has been known for some time that coordinated uctuations
between specialist regions of the brain are critical for behavior
(Friston, 1994) and that the brain is an inherently dynamic organ,
capable of exible reconguration in the face of external com-
plexities (Shine and Poldrack, 2018). One manifestation of this
reconguration is that the pattern of coordinated activities among
several brain regions that are similarly activated in different tasks
may also be distinct (Hillary et al., 2011; Yang et al., 2015; Wu
et al., 2019). For instance, a neuroimaging study by Wu et al.
Downloaded from https://academic.oup.com/scan/article/19/1/nsae001/7510586 by guest on 07 February 2024

2 Social Cognitive and Affective Neuroscience, 2024, Vol. 19, No. 1
(2019) examined the neural correlates of language control and
cognitive control and found that language control recruited more
connections from frontal to subcortical areas compared with
general cognitive control, demonstrating a recongurable brain
network.
However, previous studies exploring the neural mechanisms
underlying cognitive and emotional conict control have typi-
cally been limited to the relative activation between conditions.
Most research has not directly tested the patterns of interaction
between various key brain regions and has failed to recognize that
emotional conict control may have a reconguring effect on the
domain-general conict control brain network. Thus, it is unclear
how different types of conict modulate patterns of functional
connectivity between brain areas to support different conict con-
trols in the domain-general conict control brain network, which
is co-activated by cognitive conict control and emotional conict
control or whether there is reconguration of brain networks for
domain-general conict control that are achieved in connective
ways.
Present study
We administered two tasks involving cognitive and emotional
conict control to a group of individuals. During the tasks, we used
functional magnetic resonance imaging (fMRI) to examine the
participants’ brain activities. For the cognitive task, we adopted
an arrow version of the anker task which required participants
to respond to the direction of the middle arrow while ignoring
anking arrows (see Cognitive Conict Task section for details).
We designed an emotional conict task to be a contextualized
paradigm in which emotional conicts were triggered by emotions
themselves. In the task, participants were required to judge the
emotional valence or the opposite emotional valence of words as
indicated by an accompanying cue (see Emotional Conict Task
section for details). In choosing the two tasks, our aim was to
examine the neural overlap between the two conict tasks in
order to dene the domain-general conict control network for
subsequent analyses.
To model the causal interactions (i.e. effective connectivity)
of brain regions for cognitive functions, we adopt extended uni-
ed structural equation modeling (euSEM; Gates et al., 2011).
The euSEM method has been widely used in several studies (e.g.
Hillary et al., 2011; Yang et al., 2015; Younger et al., 2017; Wu et al.,
2019) and is a exible and efcient way of analyzing data based
on SEM. The method also allows for exploratory data analyses
without prior theoretical assumptions.
We rst conducted activation analyses to determine the pre-
cise brain regions that were critical for emotional and cognitive
conict control. We then extracted and modeled the time course
of each critical region using euSEM for both tasks and obtained
connectivity maps for emotional and cognitive conict control.
Wilcoxon signed-rank tests were performed to examine differ-
ences in the strength of shared connections for the two types
of conict control. These analyses were then used to examine
the specic connectivity patterns of two conict controls. We
hypothesize that the connectivity pattern will reveal that emo-
tional conict control recongures the domain-general conict
control network in connective ways.
Method
Participants
Thirty participants were recruited from Liaoning Normal Uni-
versity in China, and three of them were excluded due to poor
performance in the two experimental tasks (accuracy<70%).
Thus, the participants in the data analyses included 27 individ-
uals (14 females, 13 males, M=21.93 ±2.35 years), exceeding the
minimum sample size of 24 calculated by G*power 3.1.9.7. Param-
eters were set as follows: F-tests >analysis of variance (ANOVA):
repeated measures, within factors, medium effect size f=0.25,1α
error probability=0.05, correlation among repeat measures=0.5,
power (1 −β error probability) =0.8, number of groups=2, number
of measurements =4, and nonsphericity correct ∈ = 1. All par-
ticipants were right-handed, had normal or corrected-to-normal
vision and had no history of neurological or psychological dis-
orders. Participants were required to sign an informed consent
form before taking part in the study and were given a modest
monetary remuneration after their participation. The study was
approved by the Ethics Review Committee at Liaoning Normal
University.
Materials and procedure
The experimental procedure involved two tasks: the arrow ver-
sion of the anker task, which measured cognitive conict con-
trol, and an emotional valence conict task, which measured
emotional conict control. The order of the two tasks was coun-
terbalanced across participants. Participants completed practice
trials prior to beginning the formal experiment at which time,
they proceeded to a separate fMRI laboratory where the testing
took place.
Cognitive conict task
The anker task (Gibson, 1969; Eriksen and Eriksen, 1974) is a clas-
sic experiment thought to induce and measure cognitive conict
control (Zavala et al., 2013; Imburgio et al., 2020). We administered
an arrow version of anker task in which each trial started with
a white xation point (•) on a black background for 500 ms (see
Figure 1). The xation point then disappeared, and ve arrows
were displayed. Upon seeing the arrows, participants responded
to the direction of the middle arrow by pressing the appropriate
button on the fMRI keyboard as quickly and accurately as possible
(i.e. pressing ‘2’ with their left thumb to answer ‘left’ and ‘3’ with
their right thumb to answer ‘right’). The arrows were displayed for
3000 ms or until participants responded, at which time, the screen
went blank until the completion of the 3000 ms. Another blank
screen randomly appeared for 1000ms to 3000 ms before the next
trial began. Half of the trials were congruent (e.g. ««< or »»>) and
thus represented non-conict conditions, and the other half were
incongruent (e.g. «>« or »<»), representing conict conditions. The
order of the trials was randomly presented for each participant.
There were 3 warm-up trials and 72 experimental trials, including
36 non-conict trials and 36 conict trials. The number of trials in
the anker task was set to match the number of trials in the emo-
tional conict task (i.e. 36 non-conict trials of positive words and
36 conict trials of positive words, specically mentioned in Emo-
tional Conict Task section), which were ultimately included in
the analysis. The entire task lasted for 7 min.
Emotional conict task
Emotional conict control typically has been studied in previous
research using an emotional face-word Stroop task in which par-
ticipants make emotional valence judgments about faces while
ignoring the valence of words or vice versa (Etkin et al., 2006;
1For the ANOVA test, Cohen suggested the effect sizes of ‘small,’ ‘medium’
and ‘large’ as 0.1, 0.25 and 0.4, respectively (see Kang, 2021).
Downloaded from https://academic.oup.com/scan/article/19/1/nsae001/7510586 by guest on 07 February 2024

T. Guo et al. 3
Fig. 1. Procedure of the anker task.
Note: Each trial comprised a 500 ms xation point, followed by the presentation of a stimulus lasting up to 3000ms and then a random blank screen duration
ranging from 1000 ms to 3000 ms. During the presentation of the stimulus, participants were instructed to respond as quickly and accurately as possible to
indicate the direction of the middle arrow. Upon the participant’s button press, the stimulus disappeared, and a blank screen was presented to maintain the
3000 ms duration.
Fig. 2. Procedure of the emotional conict task.
Note: The example word (i.e. the Chinese word ‘快乐’) means ‘happy’. Each trial comprised a 500ms xation point, followed by the presentation of a stimulus
lasting up to 3000 ms, and then a random blank screen duration ranging from 1000ms to 3000 ms. During the presentation of the stimulus, participants were
instructed to respond to the valence of the presented words as quickly and accurately as possible. Upon the participant’s button press, the stimulus disappeared,
and a blank screen was presented to maintain the 3000ms duration.
Rey et al., 2014; Jamieson et al., 2023). Unlike the classic paradigm
which only involves dimension ignoring, we designed an emo-
tional conict task involving valence processing to induce more
contextualized emotional conicts. In the task, participants are
required to judge the emotional valence or the opposite emotional
valence of words as indicated by an accompanying cue.
In the task, participants saw individually presented Chinese
words that were accompanied by either a +or −. The +symbol
indicated a non-conict situation in which participants were to
provide the same emotional valence of the word by pressing
the appropriate button on the fMRI keyboard (i.e. pressing ‘2’
with their left thumb to answer ‘positive’ and ‘3’ with their right
thumb to answer ‘negative’). The − symbol represented a conict
condition where participants were to respond with the opposite
emotional valence. This conict condition induces a certain level
of emotional arousal conict, which arises from the disparity
between words participants saw and the valence with which they
needed to respond. Response keys were counterbalanced across
participants. Word materials consisted of 18 positive words, 18
negative words and 36 neutral words2 (for word list see Supple-
mentary Material). Moreover, ‘+ positive words’ represented a
pure non-conict condition, while ‘− positive words’ cause both
stimulus and response conict that was in line with the incon-
gruent condition of the anker task. While ‘− negative words’
cause response conict but no stimulus conict, and ‘+ nega-
tive words’ pairs cause stimulus conict but no response conict.
Hence, the conict effect of positive words was only compared
with that of the anker task. Therefore, only positive words were
included in the data analyses, while negative words and neutral
words were not relevant to the goal of the current study and were
2The 36 neutral words included 18 neutral words that were articially
characterized as positive words and 18 neutral words that were articially char-
acterized as negative words. Participants were familiar with them before the
formal experiment. For the neutral words, they performed the same task with
emotional words, but no emotion was actually involved. Thus, it was considered
a control condition and not included in the analyses.
therefore considered control trials and did not form part of the
analyses.
The procedure of the emotional conict task is shown in
Figure 2. Each trial started with a xation point (•) on a black back-
ground screen for 500 ms, followed by the visual presentation of
a target word. The word was displayed for 3000 ms or until par-
ticipants responded, at which time, the screen went blank until
the completion of the 3000ms. Another blank screen randomly
appeared for 1000 ms to 3000ms before the next trial began. A
xation point then appeared which indicated the start of a new
trial. Prior to the formal experiment, participants familiarized
themselves with the words’ valence through a reading list of the
words and performed a practice set of trials. The procedure of the
practice experiment was the same as the formal experiment, but
an additional feedback screen was presented after participants’
response to indicate whether their response was correct or not.
This feedback screen was intended to assist them in better under-
standing the task rules and was not present during the formal
experiment. The practice experiment consisted of 16 trials, and
these words were not included in the formal experiment. After
completing the practice experiments, all participants reported
that they had mastered the experimental procedure and the task
rules.
For the selection of word materials, we asked a separate age-
matched group of 24 Chinese speakers to judge their familiarity
with 182 Chinese words and their emotional arousal. Both ratings
were provided based on ve-point scales: for lexical familiar-
ity, ‘1=very unfamiliar, ‘5’ =very familiar; for emotional arousal,
‘1’ =very mild, ‘5’=very strong. The ratings showed that their
familiarity with the words was not distinct from one another: pos-
itive words (M=4.89, SD±0.16) and corresponding neutral words
(M=4.83, SD ±0.14), t(1,17) =1.220, P=0.238; negative words
(M=4.80, SD ±0.17) and corresponding neutral words (M=4.78,
SD ±0.09), t(1,17) =0.487, P=0.638; positive words and nega-
tive words, t(1,17) =1.424, P=0.172. The analyses of emotional
arousal demonstrated signicant differences between positive
Downloaded from https://academic.oup.com/scan/article/19/1/nsae001/7510586 by guest on 07 February 2024

4 Social Cognitive and Affective Neuroscience, 2024, Vol. 19, No. 1
words (M=4.42, SD ±0.22) and corresponding neutral words
(M=2.18 SD ±0.32), t(1,17) =24.179, P< 0.001, and between neg-
ative words (M=4.43, SD ±0.20) and corresponding neutral words
(M=2.08, SD ±0.28), t(1,17) =29.822, P< 0.001. There were no sig-
nicant differences found between positive words and negative
words, t(1,17) =0.132, p=0.896. Based on the analyses of lexical
familiarity and emotional arousal, we removed 110 words, leav-
ing 72 experimental words which included 18 positive words (e.g.
‘快乐’, ‘happy’) and 18 negative words (e.g. ‘悲伤’, ‘sad’); and 36
neutral words. These 36 neutral words consist of 18 pairs with
opposite meanings at the semantic level, such as ‘big’ and ‘small’,
‘long’ and ‘short’. Among them, ‘big’ and ‘long’ were articially
categorized as ‘positive words’, while ‘small’ and ‘short’ were
categorized as ‘negative words’. It is essential to note that this
‘positive/negative’ was different from the ‘positive/negative’ of the
emotional words and could be understood as analogous to posi-
tive numbers and negative numbers in mathematics. Additionally,
the lexical category of Chinese words was not xed. For instance,
快乐 (happy) could be treated as a noun, adjective or verb. The
task was divided into 4 equal blocks, each containing 3 practice
and 72 experimental trials. Each block took ∼min to complete. In
total, there were 288 trials: 36 non-conict trials of positive words
and 36 conict trials of positive words were included in the analy-
ses, and the remaining 72 trials of negative words and 144 trials of
neutral words were used as control trials and were not included
in the analyses. Trial order was randomized across participants.
Data acquisition, preprocessing and analyses
Data acquisition
Whole-brain image data were acquired by a GE Discovery MR750
3 T scanner. During scanning, participants were laid down
with their heads secured with sponges to minimize the head
motion. Functional scans were obtained using a T2*-weighted
gradient echo planar imaging (EPI) sequence. The following
scan parameters for functional images were used: slice thick-
ness =2 mm, sequential acquisition=33 axial slices, repetition
time (TR)=2000 ms, echo time (TE) =30 ms, ip angle =90∘,
image matrix=64 × 64, eld of view (FOV) =224× 224 mm and
voxel size =3.5 ×3.5 × 4.2 mm. Each functional scanning ses-
sion contained 215 time points with 5 sessions. Structural
images were collected using a three dimensional T1-weighted
magnetization prepared rapid gradient echo sequence to co-
register with the functional images using the following param-
eters: TR =6.652ms, TE =2.928 ms, ip angle=12∘, sequential
acquisition =192 slices, slice thickness =1 mm, spacing between
slices =1 mm, image matrix =256× 256, FOV =256 ×256 mm and
voxel size =1 × 1 × 1 mm.
Data preprocessing
Preprocessing of the fMRI data was performed by Data Process-
ing & Analysis for Brain Imaging (Yan et al., 2016). All of the EPI
Digital Imaging and Communications in Medicine data were con-
verted to Neuroimaging Informatics Technology Initiative format,
and the rst ve volumes of each session were discarded because
of T1 relaxation artifacts. Second, slice timing correction was per-
formed for the remaining images using the middle slice in time as
the reference slice and the images were realigned to the rst vol-
ume for head motion correction. Third, the structural images of
each participant were co-registered with mean functional images
and normalized to the Montreal Neurological Institute template.
Fourth, all voxels were resampled to 3 mm cubic voxels. Fifth, all
functional volumes were spatially smoothed using an isotropic
Gaussian kernel with a 6 mm full width at half maximum.
Behavioral and neuroimaging data analyses
The accuracy and reaction times of each participant were ana-
lyzed using a two-way repeated measure ANOVA in Statistical
Package for the Social Sciences 24. For the RT results, we removed
incorrect responses, standardized the RT values (e.g. z-scores) and
sequentially removed trials that differed by more than ±2.5 SD
from their mean RT for each participant. In doing so, this excluded
4.1% of the total data.
Brain imaging data were separately analyzed for the emo-
tional conict task and cognitive conict task using Statistical
Parametric Mapping (version 8; Wellcome Department of Cogni-
tive Neurology, London, UK, http://www.l.ion.ucl.ac.uk/spm) in
MATLAB. For each participant, a general linear model was used
to estimate the effects of the experimental conditions (i.e. con-
ict and non-conict) at the voxel-based level, with a reference
delta function of stimuli, which was convolved with a canonical
hemodynamic response function. Erroneous trials were modeled
together as a regressor of no interest and were excluded from
the analyses. The data were high-pass-ltered at 128 Hz. At the
individual level, we dened the contrast of conict condition vs
non-conict condition in each task with the conict effect (Kanske
and Kotz, 2011; Zinchenko et al., 2015). Then, for each task, a
one-sample t-test was performed on all contrast images to obtain
brain activations at the group level.
To construct a network with the same brain regions for emo-
tional conict control and cognitive conict control [i.e. to iden-
tify regions of interest (ROI) of the domain-general conict con-
trol brain network], a conjunction analysis (conjunction null
hypothesis, i.e. only voxels were reported as active if they were
signicant for the conict vs non-conict conditions in both
tasks) was performed at the group level with a family-wise error
(FWE)–corrected threshold of P< 0.05 to reject the voxels that
were not activated in either of the two tasks (Friston et al., 2005;
Nichols et al., 2005). The average time series of all the voxels in
each ROI was extracted by the Resting-State fMRI Data Analysis
Toolkit plus (Jia et al., 2019) and then used in the subsequent effec-
tive connectivity analysis (see Effective Connectivity Analyses
section).
In addition, the regions activated by conict effects of emo-
tional conict task and anker task were shown in Tables S1 and
S2 (see Supplementary Material), respectively.
Effective connectivity analyses
SEM and dynamic causal modeling (DCM) are commonly used
to construct the size and direction of the interaction between
two brain regions. Among them, DCM examines changes in node
activities, which is the sum of direct and indirect changes in
node activities caused by external stimuli and a method of model
matching data. As such, if there are too many ROIs, the research
difculty will be greatly increased. Considering the number of
ROIs in this study, SEM was chosen. SEM denes the connection
direction, establishes the model reecting the variable relation-
ship and then obtains the best-tting model by adjusting the
connection strength. The euSEM is a novel analysis based on SEM.
Traditionally, SEM provides a model of the contemporaneous rela-
tionships between ROIs and assumes that the observations are
independent. However, this is not the case for fMRI time series
since the measured fMRI signals are temporally correlated. To
address this shortcoming, Kim et al. (2007) developed the uSEM
that combines lagged (sequentially) relationships with the con-
temporaneous relationship of a conventional SEM via a multi-
variate autoregressive model. This analytical approach works well
Downloaded from https://academic.oup.com/scan/article/19/1/nsae001/7510586 by guest on 07 February 2024

T. Guo et al. 5
Fig. 3. Accuracy (A) and RTs (B) for emotional conict task and anker task.
Note: Solid lines reect medians and dotted lines represent 75% and 25% quartiles; Emo=emotional conict task, FLK=anker task, Nc =non-conict condition,
Cc =conict condition.
on fMRI data from blocked designs (Kim et al., 2007). The euSEM
builds on the uSEM and further considers task and bilinear effects
(i.e. how the relationship between two nodes changes in the pres-
ence of the task) to accommodate data from event-related fMRI
studies (Hillary et al., 2011; Gates and Molenaar, 2012; Yang et al.,
2015).
The procedure for using the euSEM is similar to that of Wu
et al. (2019). To investigate how brain networks are involved in
emotional conict control or cognitive conict control, euSEM
assessed the connectivity patterns of ROIs (which can be rep-
resented by a connectivity matrix or map) in the two tasks
separately. Using the Group Iterative Multiple Model Estima-
tion (GIMME; Gates and Molenaar, 2012), the selection of euSEM
models can be done automatically, which has been shown to
outperform other methods that attempt to model the presence
of directed connections among nodes at the group and individ-
ual levels. We implemented GIMME through the following steps.
The model selection procedure began with the use of Lagrange
multiplier equivalents (i.e. modication indices; Sörbom, 1989)
to determine which effects (including connections among ROIs
and for the euSEM and also the direct and bilinear experimen-
tal onset terms), if free, optimize the improved model to t most
individuals (more than 75%3). Next, the model was pruned by
eliminating connections that were no longer signicant for 75%
of the group after the release of the connections. Then, these
connections at the group level were freed and estimated in a semi-
conrmatory manner at the individual level. Finally, the conrma-
tory model was tted by eliminating individual-level connections
that became nonsignicant after releasing other individual-level
connections, and the model was pruned. Model t parameters
found to demonstrate that reliability were chosen a priori, so
that four criteria were satised in the nal model: conrmatory
t index (CFI) >0.90; nonnormed t index (NNFI) >0.90; standard-
ized root mean square residual (SRMR) <0.05; and root mean
square error of approximation (RMSEA) <0.08 (Gates et al., 2011.
The distinctions of the connectivity patterns were examined by a
permutation test.
3This is consistent in empirical and simulated studies examining the like-
lihood of detecting a true effect should it exist in a given sample (Hillary et al.,
2011; Gates and Molenaar, 2012).
Connection strength and hubs
The connectivity maps for emotional and cognitive conict con-
trol have shared connections. Considering the existence of out-
liers, we used the Wilcoxon sign-rank test, a non-parametric
approach which permitted us to check whether the strength of
these shared connections differed between the two tasks.
In addition to the connection from one region to another, it
is also important to identify the hubs (i.e. detected cores) of
the networks. The optimal core–periphery subdivision algorithm
(Rubinov et al., 2015; Wu et al., 2019; Yuan et al., 2021), which
divides the network into a core group and a periphery group in
such a way that the number of edges within the core group is
maximized and the number of edges within the periphery group
is minimized, was used to determine the hubs. The coreness (Q)
is reported to quantify the merit of the optimal core–periphery
subdivision. The Brain Connectivity Toolbox (Rubinov and Sporns,
2010) was used for the detection of core–periphery structure.
Results
Behavioral results
Figure 3 displays the results of the accuracy and RT analyses. For
accuracy, there was a signicant main effect of task, indicating
that participants’ accuracy in the anker task (M=0.997 ±0.002)
was higher than in the emotional conict task (M=0.979 ±0.004),
F=20.46, P< 0.001. There was also a main effect of conict
condition in which accuracy was higher in the non-conict
condition (M=0.993 ±0.003) compared to the conict condition
(M=0.982 ±0.003), F=8.87, P=0.006. Furthermore, an inter-
action between task and conict, F=6.46, P=0.017, revealed
that the conict effect in the emotional conict task (F=10.21,
P=0.004) was signicantly larger than in the anker task (F=0.39,
P=0.537).
The RT results revealed a signicant main effect of task,
showing that participants responded faster in the anker
task (M=558 ±17 ms) than in the emotional conict task
(M=935 ±32 ms), F=134.45, P< 0.001. Additionally, a main effect
of conict condition was also observed, with responses being
faster in non-conict conditions (M=656 ±19 ms) compared to
conict conditions (M=837 ±21 ms), F=351.58, P<0.001. More-
over, an interaction between task and conict was found,
F=61.42, P< 0.001, suggesting that the conict effect in the
Downloaded from https://academic.oup.com/scan/article/19/1/nsae001/7510586 by guest on 07 February 2024

6 Social Cognitive and Affective Neuroscience, 2024, Vol. 19, No. 1
Fig. 4. Activation maps for brain regions of emotional conict control, cognitive conict control and ROIs.
Note: Activation regions are averaged across 27 participants and mapped onto a standard ICBM152 brain template. Thresholds for the whole-brain and
conjunction analyses are p< 0.05 (FWE corrected); only clusters with more than 10 continuous voxels are shown; L=left, R =right.
Table 1. ROIs dened as nodes of the network
Regions MNI (x, y, z) BA t k
R/L pre-SMA 3 15 48 32 7.43 78
L dPMC −27 0 51 6 6.99 47
L pIPS −24 −60 42 7 6.82 63
R IOG 30 −90 −9 18 6.42 29
R dPMC 30 3 57 6 6.29 17
L aIPS −36 −39 42 40 6.06 13
Note: L =left, R =right. MNI represents the Montreal Neurological Institute
coordinates and BA represents the Brodman Area.
emotional conict task (F=241.48, P< 0.001) was signicantly
larger than in the anker task (F=119.78, P< 0.001).
Identication of ROIs and brain networks
The conjunction analysis of brain activation of the two types
of conict control was performed to identify ROIs. As shown in
Figure 4, thresholds for the whole-brain and conjunction analyses
are P< 0.05, FWE corrected, and clusters with more than 10 con-
tinuous voxels above the threshold were taken as ROIs, including
the pre-SMA, left and right dorsal premotor cortices (dPMCs), the
left posterior intraparietal sulcus (pIPS), the left anterior IPS (aIPS),
and the right inferior occipital gyrus (IOG). These six clusters were
selected as ROIs and dened as nodes of the network (see details
in Table 1).
Connectivity maps and hubs
Directed connectivity maps and connection strength matrixes
were generated by the euSEM for emotional conict control and
cognitive conict control. The maps revealed excellent ts to
the data for all participants: in the emotional conict task,
CFI =0.986, NNFI=0.959, SRMR =0.016 and RMSEA=0.049 and in
the cognitive conict task, CFI=0.985, NNFI =0.957, SRMR=0.035
and RMSEA =0.05. According to the results of the permutation
test, the two maps were signicantly distinct from each other
(P< 0.001).
The connectivity map of the emotional conict control
revealed a well-interconnected network and the optimal core–
periphery analysis showed that the pre-SMA was the core of the
network (Q=0.571) (see Figure 5A). The pre-SMA received inu-
ence from the left dPMC [connection strength (𝛽)=1.08] and pro-
jected information to the right dPMC (𝛽 = 0.78). Moreover, neural
signals from the left dPMC also owed along the left aIPS (𝛽= 0.60)
and the left pIPS (𝛽 = 0.60) to the right IOG (𝛽 = 0.66). The left pIPS
then sent signals to the pre-SMA (𝛽 = 0.64) as well.
The connectivity map of cognitive conict control showed both
similar and distinct patterns as the connective map of emotional
conict control (see Figure 5B). The similarity was that neural sig-
nals from the left dPMC owed along the left aIPS (𝛽 = 0.50) and
the left pIPS (𝛽 = 0.56) to the right IOG (𝛽 = 0.32). Differentially, no
core was found in the cognitive conict control network, and infor-
mation owed along the pre-SMA, the left dPMC, and the right
dPMC, forming neural circuits: from the pre-SMA to the left dPMC
(𝛽 = 0.39), from the left dPMC to the right dPMC (𝛽 = 0.30), and
from the right dPMC to the pre-SMA (𝛽 = 0.41). In addition, the
pre-SMA exerted inuence on the right IOG (𝛽 = 0.23), while the
left aIPS had an effect on the right dPMC (𝛽 = 0.28). The connec-
tion strength matrixes of emotional conict control and cognitive
conict control were shown in Figure 6.
Shared connections
The two maps shared three connections: left dPMC → left aIPS
→ left pIPS → right IOG. As shown in Figure 7, the strength of
Downloaded from https://academic.oup.com/scan/article/19/1/nsae001/7510586 by guest on 07 February 2024

T. Guo et al. 7
Fig. 5. Brain connectivity maps for emotional (A) and cognitive conict control (B).
Note: The standard ICBM152 brain template was used for brain connectivity maps. All lines represent contemporaneous relations after controlling for time-lagged
relations and the modulating effects. Solid lines indicate shared connections, and dashed lines refer to specic connections. Numbers represent connection
strengths. The pre-SMA node was the core of the Emotional Conict Control network. L=left, R =right.
Fig. 6. Connection strength matrixes of emotional (A) and cognitive conict control (B).
Note: L =left, R =right.
the shared connection from the left dPMC to the left aIPS and
from the left aIPS to the left pIPS did not show a signicant dif-
ference, Ps> 0.15. The strength of the shared connection from
the left pIPS to the right IOG was signicantly distinct, such that
stronger connections were found in the emotional conict con-
trol map compared to the cognitive conict control map, z=3.676,
P< 0.001.
Discussion
In the present study, we used fMRI to examine the neural basis
of emotional conict control and cognitive conict control. We
found that the multitasking-related activity in the pre-SMA,
bilateral dPMCs, the left pIPS, the left aIPS, and the right IOG
during both types of conict control represented common sub-
processes in controlling the efcient performance of the two
tasks. Using euSEM, we explored the connectivity patterns of
these critical brain regions separately for two conict controls.
The results showed connectivity maps for emotional conict
control and cognitive conict control and both showed excel-
lent interconnectivity in the domain-general conict control net-
work (i.e. the overlapping regions of the two). The results also
revealed that emotional conict control recongured the brain
network of domain-general conict control by modulating the
connectivity pattern. In the next subsections, we will elaborate
on this reconguration and the similarities and differences of
Downloaded from https://academic.oup.com/scan/article/19/1/nsae001/7510586 by guest on 07 February 2024

8 Social Cognitive and Affective Neuroscience, 2024, Vol. 19, No. 1
Fig. 7. Strength differences between shared connections for emotional and cognitive conict control.
Note: L =left, R =right. ***P < 0.001.
connectivity patterns between emotional and cognitive conict
control.
The results showed that emotional and cognitive conict con-
trol shared connectivity such that the left dPMC owed along
the left aIPS to the left pIPS. The left dPMC cluster obtained in
our conjunction analysis overlapped with the presumed location
of the human frontal eye eld (FEF; Paus, 1996; Zu Eulenburg
et al., 2012). The FEF is thought to facilitate visual target detection
(Grosbras and Paus, 2003) by biasing perception through atten-
tional top-down signals (Corbetta and Shulman, 2002; Langner
et al., 2012). The IPS, between the inferior parietal lobule and supe-
rior parietal lobule, is a key region for attention control (Geng and
Mangun, 2009; Gillebert et al., 2011) and has also been reported
to be involved in conict processing (Friedman-Hill et al., 2003;
Luks et al., 2007). Additionally, the aIPS is suggested to be related
to top-down reorienting of attention when responding to spatially
incongruent targets (Cieslik et al., 2010). The pIPS has been found
to facilitate target discrimination after a conict trial by direct-
ing attention to features of task-relevant stimuli (Soutschek et al.,
2013). The causal role of the pIPS in target discrimination has
been supported in a study employing both fMRI and transcranial
magnetic stimulation (TMS) (Capotosto et al., 2013).
The IPS as a neural mechanism for attentional control has also
been reported in conict studies. For instance, Luks et al. (2007)
found that the IPS was related to preparatory attentional allo-
cation during conict control, as shown by increased IPS activity
during the informative cue period and subsequent decreased IPS
activity during conict-related target processing. In the present
study, the effective connectivity of these regions in both types of
conict control may represent the mechanism of attentional con-
trol. That is, conict conditions served as salient events which the
left dPMC detected, and then the incongruent signal was trans-
mitted to the left aIPS for attention reorienting, where the left
pIPS directed attention to features of task-related stimuli to facili-
tate target recognition. This result is in line with previous research
that considered FEF and IPS to be part of the dorsal attention
network (DAN), a network that facilitates top-down control of
attention for both voluntary and goal-directed behavior (Corbetta
and Shulman, 2002; Corbetta et al., 2008).
In addition to the brain regions discussed earlier, in the present
study, attentional control was also related to the right IOG, as
shown by the effective connectivity from the left pIPS to the
right IOG in both emotional and cognitive conict control. The
IOG has been found in many studies examining conict reso-
lution (Chechko et al., 2012; Xu et al., 2016) but was usually
seen as an independent visual network (Yeo et al., 2011). The
IOG also plays a key role in attentional modulations (Grosbras
et al., 2005; Hietanen et al., 2006), and the occipital regions are
sometimes attributed to the DAN (Benedek et al., 2016). How-
ever, our results showed that the connection strength from the
left pIPS to the right IOG was signicantly stronger in emotional
conict control than in cognitive conict control. It is possi-
ble that either the more complex stimuli of emotional conict
have higher demands for attention control or that they may be
related to the modulation of emotional involvement on atten-
tion control. Studies have shown that individuals more readily
pay attention to emotional than neutral stimuli (Richards and
Blanchette, 2004) and that emotional material biases attentional
resource deployment, producing an exacerbated attentional blink
(Yiend, 2010).
In the emotional conict control brain network, the pre-SMA
was found to be the core of the network, received inuence
from the attention network (i.e. the left dPMC and the left pIPS),
and then projected information to the right dPMC. Several meta-
analyses have shown that the pre-SMA, dorsal to the anterior
cingulate cortex, is a critical brain region involved in conict con-
trol across various conict paradigms (Xu et al., 2016). This region
is considered to be associated with conict monitoring, in which
signals are sent to other regions when a conict is identied (Gara-
van et al., 2003; Iannaccone et al., 2015). The dPMC, adjacent to the
posterior primary motor cortex, is also involved in conict control.
TMS studies have reported that stimulation of dPMCs promotes
response selection, an important part of conict control (Koski
et al., 2005; Chambers et al., 2007;Duque et al., 2012). For instance,
Chambers et al. (2007) reported a functional separation of dPMC
and prefrontal cortex, in that the stimulation of the right dPMC
facilitated execution but had no effect on inhibition and the stim-
ulation of the right PFC impaired inhibition. In the present study,
for emotional conict control, the pre-SMA may be responsible
for the detection of conicting information, such that, under the
direction of the attention network, it monitored interference that
was triggered by conicting emotional valence and then signaled
other areas, such as the right dPMC, to implement subsequent
conict control and to achieve the response selection (Chambers
et al., 2007).
Moreover, in the emotional conict control brain network, the
highly efcient connection between the prefrontal and parietal
regions was consistent with previous studies in which the network
consisting of prefrontal and parietal regions was often activated in
emotional control tasks in which participants are explicitly asked
to regulate their emotions (Buhle et al., 2014; Kohn et al., 2014).
Emotional regulation refers to intentionally generating, enhanc-
ing, reducing or stopping a given emotion (Langner et al., 2018).
Emotional conict control in the present study can also be viewed
Downloaded from https://academic.oup.com/scan/article/19/1/nsae001/7510586 by guest on 07 February 2024

T. Guo et al. 9
as a form of emotional regulation (i.e. participants under the con-
ict condition were asked to change words’ emotional valence
to the opposite valence), thus requiring a favorable interaction
between the frontal and parietal cortices to facilitate the regula-
tion of emotion. The frontoparietal network consisting of the IPS
and the prefrontal cortex have been demonstrated in neurocog-
nitive models as a crucial element in the perception of emotional
stimuli (Pourtois et al., 2013; Fan et al., 2018).
For the cognitive conict control brain network, the right dPMC
received information from the attention network (i.e. the left
dPMC and the left aIPS), and then information owed along the
pre-SMA, the left dPMC, and back to the right dPMC. Studies have
shown that the dPMC is associated with the selection of correct
responses in conict control (Koski et al., 2005; Chambers et al.,
2007; Duque et al., 2012). The prominent contribution of bilat-
eral dPMCs in the current study may have emerged because the
anker task required a greater demand for resources that support
response selection and action execution, compared with con-
ict monitoring or interference suppression. Consistent with this
nding, O’Shea et al. demonstrated that stimulation of the left
dPMC led to adaptive reorganization of the right dPMC to mediate
response selection.
Interestingly, neural signals seemed to form a premotor circuit
along the right dPMC, the pre-SMA, the left dPMC, and back to the
right dPMC. This implies that the pre-SMA was more involved in
response selection and action execution with bilateral dPMCs dur-
ing cognitive conict control. This speculation accords with previ-
ous studies in which the pre-SMA, located between the prefrontal
and motor systems, was hypothesized to participate in higher-
level functions related to executive control of motor skills, such
as resolving conicting responses (Nachev et al., 2007) and select-
ing correct responses (van Gaal et al., 2011). For instance, Nachev
et al. (2007) reported that patients with lesions in the pre-SMA
displayed an impairment for inhibiting competing motor plans
in conict situations. Moreover, gray matter density in the pre-
SMA has been found to be positively correlated with the ability
to voluntarily select the correct action when resolving conicting
responses (van Gaal et al., 2011).
Finally, to examine how emotional conict control recongures
the domain-general conict control brain network, we explored
the neural separation of two conict controls at the level of spe-
cic interaction patterns in brain regions, which adds to the
exploratory nature of this study. Nonetheless, some limitations of
our study merit mention. First, the roles of various brain regions
discussed earlier are assumptions based on previous studies and
focus only on the overlaps, and ignoring regions relevant to a spe-
cic task may cause us to lose some information. The results may
be different when considering the full network. Thus, the objec-
tive of our study was not to examine the specic role played by
each brain region in different conict controls, but to underscore
that even the domain-general conict control network involved in
different conict controls at the same time is regulated by conict
types and shows unique connectivity patterns. Second, the emo-
tional conict task in the present study only contained results for
positive words. It should be pointed out that conict control for
negative words may also display specic patterns of brain net-
work connectivity. Third, the differences in task relevance of the
distracting stimuli in the two tasks (i.e. the distracting stimuli
in the emotional conict task were task relevant, whereas the
distracting stimuli in the anker task were task irrelevant) may
also affect the neural separation of the identied connectivity.
Future work should consider other tasks that can also address the
objectives of this study (i.e. the contextualization of emotional
conict control and the representation of the cognitive conict
control) while ensuring a balance between task-relevant and task-
irrelevant stimuli across experiments. This issue and the other
limitations mentioned earlier provide a foundation for further
exploration in future studies.
Conclusion
The present study examined overlapping brain networks involved
in emotional conict control and cognitive conict control and
separately explored their effective connectivity patterns using
euSEM. Specically, the ndings showed that the distributed fron-
toparietal and occipital network that was activated across tasks
worked together in emotional conict control and cognitive con-
ict control to support the different cognitive demands of conict
control. The present study has also revealed that emotional con-
ict control recongures the brain network of domain-general
conict control in a connective way. This reconguration was
reected in the fact that emotional conict control recruited dif-
ferent communication and strengthened connectivity among the
domain-general conict control network. Together, these nd-
ings offer novel evidence for the neural relationship between
emotional conict control and cognitive conict control and pro-
vide an empirical-based elaboration on the brain network under-
pinning emotional conict control and how it recongures the
domain-general conict control network in connective ways.
Supplementary data
Supplementary data is available at SCAN online.
Data availability
The datasets generated and analyzed in this study are available
in the Open Share Freedom (OSF) repository: Liu, H. (4 February
2023). Emotional Conict Control and Cognitive Conict Control.
Retrieved from osf.io/5ecwn. This dataset has also been used in
our previous research (Guo et al., 2023).
Funding
This research was supported by Grants from General Program
of National Natural Science Foundation of China (32371089),
Liaoning Social Science Planning Fund of China (L20AYY001),
Dalian Science and Technology Star Fund of China (2020RQ055),
Youth Project of Liaoning Provincial Department of Education
(LJKQZ2021089), Research and Cooperation Projects on Social and
Economic Development of Liaoning Province (2024lslybhzkt-17),
and Liaoning Educational Science Planning Project (JG21DB306).
Conict of interest
The authors declared that they had no conict of interest with
respect to their authorship or the publication of this article.
References
Benedek, M., Jauk, E., Beaty, R.E., Fink, A., Koschutnig, K., and
Neubauer, A.C. (2016). Brain mechanisms associated with inter-
nally directed attention and self-generated thought. Scientic
Reports, 6(1), 22959.
Buhle, J.T., Silvers, J.A., Wager, T.D., et al. (2014). Cognitive reappraisal
of emotion: a meta-analysis of human neuroimaging studies.
Cerebral Cortex, 24(11), 2981–90.
Downloaded from https://academic.oup.com/scan/article/19/1/nsae001/7510586 by guest on 07 February 2024

10 Social Cognitive and Affective Neuroscience, 2024, Vol. 19, No. 1
Capotosto, P., Tosoni, A., Spadone, S., et al. (2013). Anatomical segre-
gation of visual selection mechanisms in human parietal cortex.
Journal of Neuroscience, 33(14), 6225–9.
Chambers, C.D., Bellgrove, M.A., Gould, I.C., et al. (2007). Dissocia-
ble mechanisms of cognitive control in prefrontal and premotor
cortex. Journal of Neurophysiology, 98(6), 3638–47.
Chechko, N., Kellermann, T., Zvyagintsev, M., Augustin, M., Schnei-
der, F., and Habel, U. (2012). Brain circuitries involved in seman-
tic interference by demands of emotional and non-emotional
distractors. PLoS One, 7(5), e38155.
Chen, T., Becker, B., Camilleri, J., et al. (2018). A domain-general brain
network underlying emotional and cognitive interference pro-
cessing: evidence from coordinate-based and functional connec-
tivity meta-analyses. Brain Structure & Function, 223(8), 3813–40.
Cieslik, E.C., Zilles, K., Kurth, F., Eickhoff, S.B. (2010). Dissociat-
ing bottom-up and top-down processes in a manual stimulus–
response compatibility task. Journal of Neurophysiology, 104(3),
1472–83.
Corbetta, M., Patel, G., Shulman, G.L. (2008). The reorienting system
of the human brain: from environment to theory of mind. Neuron,
58(3), 306–24.
Corbetta, M., Shulman, G.L. (2002). Control of goal-directed and
stimulus-driven attention in the brain. Nature Reviews, Neuro-
science, 3(3), 201–15.
Duque, J., Labruna, L., Verset, S., Olivier, E., Ivry, R.B. (2012). Disso-
ciating the role of prefrontal and premotor cortices in control-
ling inhibitory mechanisms during motor preparation. Journal of
Neuroscience, 32(3), 806–16.
Eriksen, B.A., Eriksen, C.W. (1974). Effects of noise letters upon the
identication of a target letter in a nonsearch task. Perception &
Psychophysics, 16(1), 143–9.
Etkin, A., Egner, T., Peraza, D.M., Kandel, E.R., Hirsch, J. (2006). Resolv-
ing emotional conict: a role for the rostral anterior cingulate
cortex in modulating activity in the amygdala. Neuron, 51(6),
871–82.
Etkin, A., Prater, K.E., Hoeft, F., Menon, V., Schatzberg, A.F. (2010).
Failure of anterior cingulate activation and connectivity with the
amygdala during implicit regulation of emotional processing in
generalized anxiety disorder. American Journal of Psychiatry, 167(5),
545–54.
Fan, C., Wan, C., Zhang, J., Jin, Z., and Li, L. (2018). Repetitive
transcranial magnetic stimulation over right intraparietal sul-
cus enhances emotional face processing in the left visual eld.
Neuroreport, 29(10), 804.
Friedman-Hill, S.R., Robertson, L.C., Desimone, R., Ungerleider, L.G.
(2003). Posterior parietal cortex and the ltering of distractors.
Proceedings of the National Academy of Sciences, 100(7), 4263–8.
Friston, K.J. (1994). Functional and effective connectivity in neu-
roimaging: a synthesis. Human Brain Mapping, 2(1–2), 56–78.
Friston, K.J., Penny, W.D., Glaser, D.E. (2005). Conjunction revisited.
NeuroImage, 25(3), 661–7.
Garavan, H., Ross, T.J., Kaufman, J., Stein, E.A. (2003). A mid-
line dissociation between error-processing and response-conict
monitoring. NeuroImage, 20(2), 1132–9.
Gates, K.M., Molenaar, P.C. (2012). Group search algorithm recovers
effective connectivity maps for individuals in homogeneous and
heterogeneous samples. NeuroImage, 63(1), 310–9.
Gates, K.M., Molenaar, P.C., Hillary, F.G., Slobounov, S. (2011).
Extended unied SEM approach for modeling event-related fMRI
data. NeuroImage, 54(2), 1151–8.
Geng, J.J., Mangun, G.R. (2009). Anterior intraparietal sulcus is sensi-
tive to bottom–up attention driven by stimulus salience. Journal
of Cognitive Neuroscience, 21(8), 1584–601.
Gibson, E.J. (1969). Principles of perceptual learning and develop-
ment.
Gilbert, S.J., Burgess, P.W. (2008). Executive function. Current Biology,
18(3), R110–4.
Gillebert, C.R., Mantini, D., Thijs, V., Sunaert, S., Dupont, P., Van-
denberghe, R. (2011). Lesion evidence for the critical role of the
intraparietal sulcus in spatial attention. Brain, 134(6), 1694–709.
Grosbras, M.H., Laird, A.R., Paus, T. (2005). Cortical regions involved in
eye movements, shifts of attention, and gaze perception. Human
Brain Mapping, 25(1), 140–54.
Grosbras, M.H., Paus, T. (2003). Transcranial magnetic stimulation of
the human frontal eye eld facilitates visual awareness. European
Journal of Neuroscience, 18(11), 3121–6.
Guo, T., Schwieter, J. W., and Liu, H. (2023). fMRI reveals overlap-
ping and non-overlapping neural bases of domain-general and
emotional conict control. Psychophysiology, e14355.
Hietanen, J.K., Nummenmaa, L., Nyman, M.J., Parkkola, R.,
H
am 
al 
ainen, H. (2006). Automatic attention orienting by social
and symbolic cues activates different neural networks: an fMRI
study. NeuroImage, 33(1), 406–13.
Hillary, F.G., Medaglia, J.D., Gates, K., et al. (2011). Examining work-
ing memory task acquisition in a disrupted neural network. Brain,
134(5), 1555–70.
Iannaccone, R., Hauser, T.U., Staempi, P., Walitza, S., Brandeis, D.,
Brem, S. (2015). Conict monitoring and error processing: new
insights from simultaneous EEG–fMRI. NeuroImage, 105, 395–407.
Imburgio, M.J., Banica, I., Hill, K.E., Weinberg, A., Foti, D., and MacNa-
mara, A. (2020). Establishing norms for error-related brain activity
during the arrow anker task among young adults. NeuroImage,
213, 116694.
Jamieson, G.A., Page, J., Evans, I.D., and Hamlin, A. (2023). Conict and
control in cortical responses to inconsistent emotional signals in
a face-word Stroop. Frontiers in Human Neuroscience, 17, 955171.
Jia, X.Z., Wang, J., Sun, H.Y., et al. (2019). RESTplus: an improved
toolkit for resting-state functional magnetic resonance imaging
data processing. Science Bulletin, 64(14), 953–4.
Kang, H. (2021). Sample size determination and power analysis using
the G*Power software. Journal of Educational Evaluation for Health
Professions, 18, 17.
Kanske, P., Kotz, S.A. (2011). Emotion triggers executive attention:
anterior cingulate cortex and amygdala responses to emotional
words in a conict task. Human Brain Mapping, 32(2), 198–208.
Kim, J., Zhu, W., Chang, L., Bentler, P.M., Ernst, T. (2007). Unied
structural equation modeling approach for the analysis of multi-
subject, multivariate functional MRI data. Human Brain Mapping,
28(2), 85–93.
Kohn, N., Eickhoff, S.B., Scheller, M., Laird, A.R., Fox, P.T., Habel, U.
(2014). Neural network of cognitive emotion regulation—an ALE
meta-analysis and MACM analysis. NeuroImage, 87, 345–55.
Koski, L., Molnar-Szakacs, I., Iacoboni, M. (2005). Exploring the contri-
butions of premotor and parietal cortex to spatial compatibility
using image-guided TMS. NeuroImage, 24(2), 296–305.
Langner, R., Kellermann, T., Eickhoff, S.B., et al. (2012). Staying respon-
sive to the world: modality-specic and -nonspecic contribu-
tions to speeded auditory, tactile, and visual stimulus detection.
Human Brain Mapping, 33(2), 398–418.
Langner, R., Leiberg, S., Hoffstaedter, F., Eickhoff, S.B. (2018). Towards
a human self-regulation system: common and distinct neural
signatures of emotional and behavioural control. Neuroscience and
Biobehavioral Reviews, 90, 400–10.
Luks, T.L., Simpson, G.V., Dale, C.L., Hough, M.G. (2007). Preparatory
allocation of attention and adjustments in conict processing.
NeuroImage, 35(2), 949–58.
Downloaded from https://academic.oup.com/scan/article/19/1/nsae001/7510586 by guest on 07 February 2024

T. Guo et al. 11
Nachev, P., Wydell, H., O’ Neill, K., Husain, M., Kennard, C. (2007).
The role of the pre-supplementary motor area in the control of
action. NeuroImage, 36, T155–63.
Nichols, T., Brett, M., Andersson, J., Wager, T., Poline, J.B. (2005). Valid
conjunction inference with the minimum statistic. NeuroImage,
25(3), 653–60.
Ochsner, K.N., Hughes, B.L., Robertson, E.R., Cooper, J.C., Gabrieli,
J.D. (2009). Neural systems supporting the control of affective and
cognitive conicts. Journal of Cognitive Neuroscience, 21, 1842–55.
Paus, T. (1996). Location and function of the human frontal eye-eld:
a selective review. Neuropsychologia, 34(6), 475–83.
Pourtois, G., Schettino, A., Vuilleumier, P. (2013). Brain mechanisms
for emotional inuences on perception and attention: what is
magic and what is not. Biological Psychology, 92(3), 492–512.
Rey, G., Desseilles, M., Favre, S., et al. (2014). Modulation of
brain response to emotional conict as a function of current
mood in bipolar disorder: preliminary ndings from a follow-up
state-based fMRI study. Psychiatry Research: Neuroimaging, 223(2),
84–93.
Richards, A., and Blanchette, I. (2004). Independent manipulation of
emotion in an emotional stroop task using classical conditioning.
Emotion, 4(3), 275.
Ridderinkhof, K.R., Ullsperger, M., Crone, E.A., Nieuwenhuis, S. (2004).
The role of the medial frontal cortex in cognitive control. Science,
306(5695), 443–7.
Rubinov, M., Sporns, O. (2010). Complex network measures of
brain connectivity: uses and interpretations. NeuroImage, 52(3),
1059–69.
Rubinov, M., Ypma, R.J., Watson, C., Bullmore, E.T. (2015). Wiring cost
and topological participation of the mouse brain connectome.
Proceedings of the National Academy of Sciences, 112(32), 10032–7.
Shine, J.M., Poldrack, R.A. (2018). Principles of dynamic network
reconguration across diverse brain states. NeuroImage, 180,
396–405.
Sörbom, D. (1989). Model modication. Psychometrika, 54(3), 371–84.
Soutschek, A., Taylor, P.C., Müller, H.J., Schubert, T. (2013). Disso-
ciable networks control conict during perception and response
selection: a transcranial magnetic stimulation study. Journal of
Neuroscience, 33(13), 5647–54.
Torres-Quesada, M., Korb, F.M., Funes, M.J., Lupiáñez, J., and
Egner, T. (2014). Comparing neural substrates of emotional vs.
non-emotional conict modulation by global control context.
Frontiers in Human Neuroscience, 8, 66.
van Gaal, S., Scholte, H.S., Lamme, V.A., Fahrenfort, J.J., Rid-
derinkhof, K.R. (2011). Pre-SMA gray-matter density predicts indi-
vidual differences in action selection in the face of conscious and
unconscious response conict. Journal of Cognitive Neuroscience,
23(2), 382–90.
Wang, L., LaBar, K.S., Smoski, M., et al. (2008). Prefrontal mecha-
nisms for executive control over emotional distraction are altered
in major depression. Psychiatry Research: Neuroimaging, 163(2),
143–55.
Wu, J., Yang, J., Chen, M., et al. (2019). Brain network reconguration
for language and domain-general cognitive control in bilinguals.
NeuroImage, 199, 454–65.
Xing, Z., Guo, T., Ren, L., Schwieter, J.W., and Liu, H. (2023). Spatiotem-
poral evidence uncovers differential neural activity patterns in
cognitive and affective conict control. Behavioural Brain Research,
451, 114522.
Xu, M., Xu, G., and Yang, Y. (2016). Neural systems underlying emo-
tional and non-emotional interference processing: An ALE meta-
analysis of functional neuroimaging studies. Frontiers in Behavioral
Neuroscience, 10, 220.
Yang, J., Gates, K.M., Molenaar, P., Li, P. (2015). Neural changes under-
lying successful second language word learning: an fMRI study.
Journal of Neurolinguistics, 33, 29–49.
Yan, C.G., Wang, X.D., Zuo, X.N., and Zang, Y.F. (2016). DPABI: data
processing & analysis for (resting-state) brain imaging. NeuroIn-
formatics, 14(3), 339–51.
Yeo, B.T., Krienen, F.M., Sepulcre, J., et al. (2011). The organization
of the human cerebral cortex estimated by intrinsic functional
connectivity. Journal of Neurophysiology, 106(3): 1125–1165.
Yiend, J. (2010). The effects of emotion on attention: a review of atten-
tional processing of emotional information. Cognition and Emotion,
24(1), 3–47.
Younger, J.W., Tucker-Drob, E., and Booth, J.R. (2017). Longitudinal
changes in reading network connectivity related to skill improve-
ment. NeuroImage, 158, 90–8.
Yuan, Q., Wu, J., Zhang, M., et al. (2021). Patterns and networks of
language control in bilingual language production. Brain Structure
& Function, 226, 963–77.
Zavala, B., Brittain, J.S., Jenkinson, N., et al. (2013). Subthalamic
nucleus local eld potential activity during the Eriksen anker
task reveals a novel role for theta phase during conict monitor-
ing. Journal of Neuroscience, 33(37), 14758–66.
Zinchenko, A., Kanske, P., Obermeier, C., Schröger, E., and Kotz,
S.A. (2015). Emotion and goal-directed behavior: ERP evidence
on cognitive and emotional conict. Social Cognitive & Affective
Neuroscience, 10(11), 1577–87.
Zu Eulenburg, P., Caspers, S., Roski, C., and Eickhoff, S.B. (2012). Meta-
analytical denition and functional connectivity of the human
vestibular cortex. NeuroImage, 60(1), 162–9.
Downloaded from https://academic.oup.com/scan/article/19/1/nsae001/7510586 by guest on 07 February 2024

Social Cognitive and Affective Neuroscience, 2024, 19(1), 11–11
DOI: https://doi.org/10.1093/scan/nsae001
Advance Access Publication Date: 4 January 2024
Original Research – Neuroscience
Received: 24 March 2023; Revised: 24 October 2023; Accepted: 3 January 2024
© The Author(s) 2024. Published by Oxford University Press.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which
permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
Downloaded from https://academic.oup.com/scan/article/19/1/nsae001/7510586 by guest on 07 February 2024