Content uploaded by Martial Van der Linden
Author content
All content in this area was uploaded by Martial Van der Linden on Feb 07, 2019
Content may be subject to copyright.
This article was downloaded by: [UCL Service Central des Bibliothèques]
On: 05 June 2014, At: 01:08
Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office:
Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Click for updates
Cognition and Emotion
Publication details, including instructions for authors and subscription
information:
http://www.tandfonline.com/loi/pcem20
Do emotional stimuli interfere with two
distinct components of inhibition?
Marie My Lien Rebeteza, Lucien Rochatab, Joël Billieuxc, Philippe Gayabd &
Martial Van der Lindenabe
a Cognitive Psychopathology and Neuropsychology Unit, University of
Geneva, Geneva, Switzerland
b Swiss Centre for Affective Sciences, University of Geneva, Geneva,
Switzerland
c Laboratory for Experimental Psychopathology, Psychological Sciences
Research Institute, Catholic University of Louvain, Louvain-La-Neuve,
Belgium
d Haute Ecole Pédagogique du Valais, St-Maurice, Switzerland
e Cognitive Psychopathology Unit, University of Liège, Liège, Belgium
Published online: 02 Jun 2014.
To cite this article: Marie My Lien Rebetez, Lucien Rochat, Joël Billieux, Philippe Gay & Martial Van der
Linden (2014): Do emotional stimuli interfere with two distinct components of inhibition?, Cognition and
Emotion, DOI: 10.1080/02699931.2014.922054
To link to this article: http://dx.doi.org/10.1080/02699931.2014.922054
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”)
contained in the publications on our platform. However, Taylor & Francis, our agents, and our
licensors make no representations or warranties whatsoever as to the accuracy, completeness, or
suitability for any purpose of the Content. Any opinions and views expressed in this publication
are the opinions and views of the authors, and are not the views of or endorsed by Taylor &
Francis. The accuracy of the Content should not be relied upon and should be independently
verified with primary sources of information. Taylor and Francis shall not be liable for any
losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities
whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or
arising out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any substantial
or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or
BRIEF REPORT
Do emotional stimuli interfere with two distinct
components of inhibition?
Marie My Lien Rebetez
1
, Lucien Rochat
1,2
, Joël Billieux
3
, Philippe Gay
1,2,4
, and
Martial Van der Linden
1,2,5
1
Cognitive Psychopathology and Neuropsychology Unit, University of Geneva, Geneva, Switzerland
2
Swiss Centre for Affective Sciences, University of Geneva, Geneva, Switzerland
3
Laboratory for Experimental Psychopathology, Psychological Sciences Research Institute, Catholic
University of Louvain, Louvain-La-Neuve, Belgium
4
Haute Ecole Pédagogique du Valais, St-Maurice, Switzerland
5
Cognitive Psychopathology Unit, University of Liège, Liège, Belgium
Emotions have recently been shown to interfere with the efficacy of inhibitory control. However,
understanding their impact requires taking into account that inhibition is not a unitary construct, but
consists of distinct functions underlain by specific mechanisms. In this study, 88 participants
performed two emotional versions of classic laboratory tasks designed to assess (1) the ability to inhibit
a prepotent response (a stop-signal task using faces with different emotional expressions) and (2) the
capacity to resist the effect of proactive interference (PI; a recent negative task that included emotional
words). Overall results showed that emotional stimuli interfered with inhibition capacities in both
tasks. Although tending in the same direction, these results suggest that different underlying
mechanisms (e.g., top-down vs. bottom-up processes) or consecutive differences in emotional
processing (e.g., different interactions with stimulus/task properties, processing stages or motivational
aspects) are at play in these two inhibition-related functions.
Keywords:Executive functions; Prepotent response inhibition; Proactive interference; Emotion.
Emotional experiences are now acknowledged to
interfere with the effectiveness of inhibitory
control, emotional stimuli capturing attention
automatically, interrupting ongoing activities
and leaving fewer resources available for effortful
control (Schimmack, 2005).
Understanding the impact of emotions on
inhibitory control, however, requires taking into
Correspondence should be addressed to: Marie My Lien Rebetez, Cognitive Psychopathology and Neuropsychology Unit, FPSE,
University of Geneva, Boulevard du Pont d’Arve, 40, CH-1205 Geneva, Switzerland. E-mail: Marie.Rebetez@unige.ch
© 2014 Taylor & Francis 1
COGNITION AND EMOTION, 2014
http://dx.doi.org/10.1080/02699931.2014.922054
Downloaded by [UCL Service Central des Bibliothèques] at 01:08 05 June 2014
account that inhibition is not a unitary construct.
Indeed, Friedman and Miyake (2004) have dem‐
onstrated by using latent variable analyses that
prepotent response inhibition (i.e., the ability to
deliberately control or suppress an automatic
response) is unrelated to resistance to proactive
interference (PI; i.e., the ability to resist the
intrusion into memory of information that was
previously relevant but has since become irrelev-
ant). One possible explanation given by Friedman
and Miyake on this separability is that prepotent
response inhibition requires effortful control (i.e.,
the active maintenance of a task’s goal), whereas
resistance to PI is more automatically driven (the
source of interference residing in memory and not
interfering with the maintenance of the task’s
goal). In this context, it is therefore necessary to
examine whether emotional experiences interfere
with the effectiveness of these different compo-
nents of inhibition in the same way.
On the one hand, a growing corpus of data
supports the concept that emotions impair the
ability to inhibit prepotent responses. First, Schulz
et al. (2007), using two comparable emotional and
non-emotional go/no-go tasks, showed that inhib-
iting a prepotent response was more difficult in
the emotional than in the non-emotional task.
Second, Verbruggen and De Houwer (2007) used
a stop-signal task (SST) in which emotional
stimuli preceded neutral targets. Results showed
that both responding to neutral targets and
stopping the response were impaired by the
previous presentation of high arousal emotional
stimuli. This interference effect has been linked to
the attentional account that emotional stimuli
attract attention away from the task, leaving fewer
resources available for its correct application
(Schimmack, 2005; Verbruggen & De Houwer,
2007). Finally, by using a different SST than those
developed by Verbruggen and De Houwer,
Herbert and Sütterlin (2011) also showed that
emotional stimuli interfered with prepotent
response inhibition. Indeed, they used an SST in
which emotional words were the target of
response inhibition. This SST thus measured the
response inhibition of emotional targets, whereas
Verbruggen and De Houwer’s SST measured the
response inhibition of neutral targets preceded by
emotional stimuli.
On the other hand, emotional stimuli have
been shown to improve the capacity to resist the
effect of PI. Levens and Phelps (2008) used a
recent negative task (RNT) that included emo-
tional words, in which previously learned informa-
tion interfered with new learning. In this
paradigm, participants were given a list of three
words followed by a probe word and had to
determine whether the probe word was included
in the list (yes-response) or not (no-response). In
no-response trials, the probe word could either
match an item from the preceding two lists [recent
no-response (RecNR) trials] or no item of any list
[non-recent no-response (NRecNR) trials]. All
trials were grouped into an emotional condition
containing both neutral and emotional words (i.e.,
three neutral and emotional words followed by
either an emotional or a neutral probe) or into a
neutral condition containing only neutral words.
When RecNR trials (interference condition) and
NRecNR trials (non-interference condition) were
compared, poorer performances were observed in
the interference condition. This suggested that the
previously yet-no-longer relevant information
interfered with the processing of the currently
relevant information (i.e., PI) and that time was
needed to resolve this conflict between relevant
and irrelevant working memory representations
(i.e., interference resolution). However, when
emotional and neutral conditions were compared,
interference was observed to decrease in emotional
probe trials. According to Levens and Phelps, a
correct response in the interference condition
required to resolve a conflict between the source
recognition signal of this response (“no”) and the
familiarity signal of the incorrect response (“yes”).
A decrease in interference therefore implied an
increase in source signal strength or a decrease in
familiarity signal strength. Thus, two hypotheses
have been put forward to explain the facilitating
effect of emotional stimuli on interference resolu-
tion: The first suggests a decrease in the familiarity
signal strength of the previously yet-no-longer
relevant information by special mechanisms for
interference resolution with emotional stimuli; and
REBETEZ ET AL.
2COGNITION AND EMOTION, 2014
Downloaded by [UCL Service Central des Bibliothèques] at 01:08 05 June 2014
the second suggests an increase in the source
signal strength of the currently relevant informa-
tion by enhanced encoding of emotional source
memory.
According to the literature, emotional experi-
ences seem, therefore, to either impair or facilitate
inhibition, depending on which inhibition-related
function is examined. These opposite effects could
be in line with the implication that different
mechanisms occur in prepotent response inhibi-
tion and in resistance to PI, as suggested by
Friedman and Miyake (2004). In this context,
the main objective of this study was to examine
the influence of emotional stimuli on both prepo-
tent response inhibition and resistance to PI.
More specifically, our first objective was to
replicate the established interference effect of
emotions on prepotent response inhibition by
using an SST that measured the response inhibi-
tion of emotional stimuli with facial expressions as
stimuli. Indeed, a distinction must be made
between an SST in which emotional stimuli
precede the inhibition of neutral targets (see
Verbruggen & De Houwer, 2007) and an SST
in which emotional stimuli are the target of
response inhibition themselves (see Herbert &
Sütterlin, 2011), these two types of SST poten-
tially involving distinct processing stages. More-
over, an interference effect of emotions on the
prepotent response inhibition of emotional stimuli
themselves was provided by using words as stimuli
(Herbert & Sütterlin, 2011). This effect should
not be restricted to verbal stimuli, however, but
should also occur when pictorial stimuli are used.
Thus, we used an SST that included faces with
different emotional expressions as the target of
response inhibition.
Our second objective was to explore the impact
of emotional stimuli on resistance to PI with a
paradigm that differed from that of Levens and
Phelps (2008) in terms of (1) distinctiveness (i.e.,
the degree to which a word stands out relative to
other words in a list), (2) time of stimuli
presentation and (3) number of stimuli. Indeed,
resistance to PI would be relatively automatically
driven (Friedman & Miyake, 2004) and thus
dependent on bottom-up processes influenced by
the environment. Such an inhibitory mechanism
would thus be particularly sensitive to properties of
the paradigm. The presentation of emotional
stimuli in mixed lists, as is the case in Levens
and Phelps’paradigm, has been shown to be more
distinctive and produce better recall by linking
separate events together (Talmi, Luk, McGarry, &
Moscovitch, 2007). Therefore, changes at that
level (i.e., in the way emotional stimuli stand out
from their surroundings) would reasonably lead to
differences in the way PI is resolved for emotional
stimuli. Moreover, Levens and Phelps suggested
that the reduction in PI for emotional stimuli was
possibly due to a decrease in the familiarity signal
and an increase in source recognition. However,
shortening the time of stimuli presentation could
lead to a decrease in source recognition (through
diminished encoding of source memory); more-
over, reducing the number of stimuli (which are
therefore repeated more frequently) could improve
the familiarity signal.
Thus, we used an RNT that included emo-
tional target sets of three sequential words with
the same valence, in which the time of stimuli
presentation was quicker and the number of
stimuli fewer than those used in the paradigm of
Levens and Phelps (2008).
METHOD
Participants and procedure
Participants comprised 88 volunteers from the
community (55 females and 30 males; three
missing values) who received no compensation
for their participation and were recruited through
advertisement and personal contacts. The mean
age of the sample was 22.78 years (SD = 2.35,
range = 18–30) and the mean number of years of
education was 14.70 (SD = 1.86, range = 9–20).
The inclusion criterion was being a fluent speaker
of French. Exclusion criteria were a reported
history of neurological or psychiatric disorders.
Participants were individually tested in a quiet
room. They signed an informed consent form
before completing tasks for which the order was
HOW EMOTION INTERFERES WITH INHIBITION
COGNITION AND EMOTION, 2014 3
Downloaded by [UCL Service Central des Bibliothèques] at 01:08 05 June 2014
counterbalanced. This study was approved by the
Ethics Committee of the University of Geneva.
Laboratory tasks
Emotional SST
This task was used to assess the interference effect of
emotions on prepotent response inhibition of emo-
tional stimuli themselves adapted from Billieux, Gay,
Rochat, & Van der Linden (2010). Stimuli consisted
of pictures of human faces with neutral, joy or angry
expressions selected from the Karolinska Directed
Expressional Faces set (Lundqvist, Flykt, & Öhman,
1998). Each face was shown for a maximum
duration of 2000 ms (the face disappearing once
the participant answered) and was preceded by a
fixation cross displayed for 500 ms. The task had
two distinct parts: an automation phase and a stop-
signal phase. In the automation phase, intended to
build a prepotent categorisation response, 48 trials
were conducted in which participants had to cate-
gorisethegenderofthefacesasquicklyaspossible
without making mistakes. The stop-signal phase was
composed of 20 practice trials followed by four
blocks of 96 trials (16 female and 16 male faces, each
with the three different emotional expressions,
presented once in each block). During this phase, a
computer-emitted tone (stop-signal) was presented
in 25% of the trials at predetermined intervals. The
length of the interval before the occurrence of the
stop-signal was 200, 240, 280 or 320 ms (two
identical intervals were never presented sequentially).
Participants were told not to respond when the stop-
signal occurred (“stop”trials) but otherwise to keep
performing the same categorisation task (“go”trials).
They received feedback at the end of each block on
the mean reaction time (RT) of the block and of the
training phase. This feedback was needed to reduce
the tendency for participants to wait until the
occurrence of the stop-signal to answer. The stop-
signal reaction time (SSRT) that reflects the latency
of the inhibitory process (see Logan, 1994)canbe
estimated from the start of the stop process (experi-
mentally controlled by the interval before the occur-
rence of the stop-signal) and its finish (inferred from
the RT distribution in the observed no-stop trials; if
responses are not stopped in n%ofthestoptrials,
the finish of the stop process is on average equal to
the nth RT distribution for the go trials). The SSRT
was thus calculated (for each interval and then
averaged) by subtracting the mean of the stop-signal
interval from the nth RT distribution for go trials.
Only the RT for correct responses was retained and
every go trial lower than 200 ms or longer than the
mean for go trials plus 2.5 SD was suppressed on a
subject-by-subject basis. Dependent variables were
the SSRT computed separately for neutral, joy and
angry conditions.
Emotional RNT
This task was designed to assess the impact of
emotional stimuli on the resolution of PI in working
memory adapted from Gay, Rochat, Billieux,
d’Acremont, & Van der Linden (2008). It was
composed of neutral, positive or negative target sets
of three sequential words drawn from a list of 16
neutral, 16 positive or 16 negative disyllabic words
that were matched for length, lexical frequency,
arousal and imagery. Each word of the target set
was presented sequentially for 750 ms, separated
from the next word by an interval of 100 ms, with a
row of number signs displayed for 400 ms after the
third word, and followed by the presentation of a
single probe word for 600 ms. Participants were
instructed to indicate, as quickly as possible without
making mistakes, whether the probe word was
presented in the last set of three words (yes-
response) or not (no-response). When the probe
word was not presented in the last trial’stargetset,
two conditions were distinguished: (1) RecNR, in
which probes occurred in the previous trial’starget
set; and (2) NRecNR, in which probes occurred
three trials before the current one. There were a total
of 240 trials divided into three blocks (one block by
emotional condition; i.e., neutral, positive or negative
words) randomised across participants, as well as two
practice trials that were not scored. In each block, 40
trials required a no-response (among which 20 trials
were the RecNR condition and 20 trials were the
NRecNR condition) and 40 trials required a yes-
response. For each condition, only the RT for
correct responses was retained and every RT lower
REBETEZ ET AL.
4COGNITION AND EMOTION, 2014
Downloaded by [UCL Service Central des Bibliothèques] at 01:08 05 June 2014
than 200 ms or longer than the mean for RT plus
2.5 SD was suppressed on a subject-by-subject basis.
Dependent variables were number of errors and RT
computed separately for RecNR and NRecNR in
each emotional condition. Since the PI index is the
more common measure used to assess PI, and in
ordertotakeintoaccountthebaseperformanceof
the participants, we also computed PI indices on the
basis of errors (errors in NRecNR condition minus
errors in RecNR condition) and RT (RT in
NRecNR condition minus RT in RecNR condition)
for each emotional condition.
RESULTS
An exploratory analysis of extreme values (box plot
of multiple variables) on measures for the SST and
the RNT revealed four participants with statistical
extreme values (i.e., greater than the upper quartile
plus three times the interquartile range). The
values for these participants were thus discarded
from the analyses.
Mean scores and standard deviations on mea-
sures for the SST and the RNT are presented in
Table 1. Because the focus here is inhibitory
control, for the sake of conciseness, we have
chosen to report only inhibition-related variables.
Partial correlations performed between the
SSRT of the SST and measures in the RecNR
condition of the RNT (removing the variability
explained by measures in the NRecNR condition)
indicated no significant relationship between
the SSRT of the SST and errors (r= .19) or RT
(r= .08) in the RNT.
Repeated measures analyses of variance (ANO-
VAs) were then performed to evaluate the effect of
emotional stimuli on task performances.
Concerning the SST, ANOVA with trial type
(neutral vs. positive vs. negative) as the within-
subject factor performed for SSRTs revealed a
significant main effect of trial type, F(2, 166) =
16.02, p<.001, g2
p¼.16. Tukey’s honestly
significant difference post hoc tests indicated that
stopping was prolonged on negative trials com-
pared with neutral trials (p<.001) and positive
trials (p<.05); stopping on positive trials was
prolonged compared with neutral trials (p<.05).
Concerning the RNT, ANOVAs with recency
effect (recent vs. non-recent) × trial type (neutral
vs. positive vs. negative) as within-subject factors
were performed separately for errors and RTs.
ANOVA computed for errors revealed a signific-
ant main effect of recency, F(1, 83) = 58.49,
p<.001, g2
p¼.41, and trial type, F(2, 166) =
8.05, p<.001, g2
p¼.09, as well as a significant
interaction effect between recency and trial type,
F(2, 166) = 4.59, p<.05, g2
p¼.05.
1
ANOVA
computed for RTs showed a significant main
effect of recency, F(1, 83) = 84.91, p<.001,
g2
p¼.51, but no effect of trial type, F(2, 166) =
1.25, g2
p¼.01, or interaction, F(2, 166) = .47,
g2
p¼.01. With regard to recency effects, errors
were higher and RTs longer on recent than on
non-recent trials. ANOVA with trial type as the
within-subject factor was then performed for the
PI index based on errors. This analysis revealed a
significant main effect of trial type, F(2, 166) =
4.49, p<.05, g2
p¼.05. As indicated by post hoc
tests, the PI index based on errors was higher
on positive trials compared with neutral trials
(p<.05) and negative trials (p<.05); there was
no difference between neutral and negative trials.
Ordinary non-parametric bootstrap analyses were
finally performed (comparison of the PI index
based on errors for positive trials with those based
on errors for neutral trials and for negative trials)
to support the robustness of these results given the
low error rates. These analyses confirmed that the
PI index based on errors for positive trials differed
significantly from those based on errors for neutral
1
In order to analyse this interaction effect, we also performed two additional ANOVAs with trial type (neutral vs.
positive vs. negative) as the within-subject factor for recent and non-recent conditions separately. Analyses revealed a
significant main effect of trial type for the recent condition, F(2, 166) = 8.82, p<.001, g2
p¼.10. As indicated by post hoc
tests, errors were higher on positive trials compared with neutral trials (p<.001) and negative trials (p<.01); there was no
difference between neutral and negative trials. No significant main effect of trial type was observed for the non-recent
condition, F(2, 166) = 1.04, g2
p¼.01.
HOW EMOTION INTERFERES WITH INHIBITION
COGNITION AND EMOTION, 2014 5
Downloaded by [UCL Service Central des Bibliothèques] at 01:08 05 June 2014
trials (95% CI: .14–1.21) and for negative trials
(95% CI: .11–1.36).
DISCUSSION
The goal of the current study was to examine the
influence of emotional stimuli on two inhibitory
functions: prepotent response inhibition and res-
istance to PI. Results showed a deleterious effect
of emotions on these two inhibitory functions.
Regarding the SST, it was found that stopping
latencies were prolonged on emotional trials com-
pared with neutral ones, providing further evidence
that emotional experiences interfere with the
effectiveness of inhibitory control by capturing
attention automatically and guiding behaviour.
More specifically, greater interference was observed
for negative valence than for positive valence, which
corroborated previous studies showing that positive
stimuli elicit less attention than negative stimuli do
(therefore producing less interference; see Baumeis-
ter, Bratslavsky, Finkenauer, & Vohs, 2001).
A motivational aspect might also account for this
finding. Indeed, observed prioritisation of the pro-
cessing of negative stimuli may be driven by the
protection of the self, as negative stimuli quickly
signal potential danger in the environment
and prepare us to face such danger by interrupting
ongoing behaviour and mental processes (Öhman &
Mineka, 2001).
In addition, the SST used in the current study
is of special interest for at least two reasons. First,
emotional stimuli were the target of response
Table 1. Descriptive statistics for laboratory tasks
Task Variable Mean (SD) Error rate (%)
SST GO RT-neutral 474.72 (53.77)
GO RT-positive 475.75 (53.08)
GO RT-negative 480.36 (52.21)
STOP errors-neutral 10.67 (5.25) 33.33
STOP errors-positive 11.43 (5.52) 35.71
STOP errors-negative 12.09 (5.50) 37.80
SSRT-neutral 156.45 (39.64)
SSRT-positive 166.63 (44.87)
SSRT-negative 176.20 (38.03)
RNT NRecNR RT-neutral 573.61 (104.44)
NRecNR RT-positive 579.25 (110.29)
NRecNR RT-negative 572.67 (100.71)
RecNR RT-neutral 611.24 (124.03)
RecNR RT-positive 623.32 (120.44)
RecNR RT-negative 609.60 (126.15)
NRecNR errors-neutral 0.83 (1.14) 4.17
NRecNR errors-positive 1.01 (1.20) 5.06
NRecNR errors-negative 1.04 (1.37) 5.18
RecNR errors-neutral 1.71 (1.74) 8.57
RecNR errors-positive 2.57 (2.25) 12.86
RecNR errors-negative 1.86 (2.03) 9.29
PI index based on RT-neutral 37.63 (61.61)
PI index based on RT-positive 44.07 (52.47)
PI index based on RT-negative 39.93 (72.80)
PI index based on errors-neutral .88 (1.81)
PI index based on errors-positive 1.56 (2.19)
PI index based on errors-negative .82 (1.79)
SST, stop-signal task; GO RT, reaction time for go response; STOP errors, errors for stop response; SSRT, stop-signal reaction time; RNT,
recent negative task; NRecNR, non-recent no-response condition; RecNR, recent no-response condition; PI, proactive interference.
REBETEZ ET AL.
6COGNITION AND EMOTION, 2014
Downloaded by [UCL Service Central des Bibliothèques] at 01:08 05 June 2014
inhibition, implying the task specifically measured
response inhibition regarding emotional stimuli.
The task used has thus to be distinguished from
the SST used by Verbruggen and De Houwer
(2007) in which emotional stimuli preceded the
inhibition of neutral targets. This distinction
among the two tasks is crucial, given that a series
of electrophysiological studies showed that distinct
processing stages are involved in the emotional
guidance of attention, as well as in the interaction
of emotion and attention (see Schupp, Flaisch,
Stockburger, & Junghöfer, 2006). For example,
paying attention to emotional stimuli (when emo-
tional stimuli are the target of response inhibition)
rather than to neutral stimuli (when neutral targets
preceded by emotional stimuli must be inhibited)
could be more effective in drawing attentional
resources (as attention is explicitly directed to
emotional stimuli), allowing overadditive effects
of the interaction between emotion and attention
(Schupp et al., 2006). Second, we used faces with
various expressions as stimuli, whereas verbal
stimuli were previously used with an SST focusing
on response inhibition of emotional stimuli
(Herbert & Sütterlin, 2011). The more specific
use of facial expressions is of particular relevance,
as faces are powerful social cues that guide
behaviour in everyday life (Yoon, Joorman, &
Gotlib, 2009) and influence the production and
regulation of affective states (Phillips, Drevets,
Rauch, & Lane, 2003).
Concerning results found with the RNT,
longer RTs and higher error rates for interference
than for non-interference trials indicate that pre-
viously yet-no-longer relevant words interfered
with the processing of currently relevant words,
replicating the PI effect (Gay et al., 2008;
Jonides & Nee, 2006; Levens & Phelps, 2008).
In particular, greater interference for positive trials
was observed, as demonstrated by a higher PI
index based on errors for positive trials relative to
that based on errors for neutral trials (and higher
error rates for recent positive trials relative to those
for recent neutral trials). As with the SST, a
motivational aspect might account for the specifi-
city of this finding for positive words. Indeed,
enhanced memory for positive stimuli could be
motivated here by the achievement or mainten-
ance of emotional well-being (Carstensen &
Mikels, 2005). Nonetheless, these results contrast
with those in the study by Levens and Phelps
(2008), who found a facilitation of both positive
and negative stimuli on the resolution of PI in
working memory. The specificity of our paradigm
could explain, at least in part, these divergent
findings. First, the distinctiveness of emotional
stimuli was controlled in our paradigm. Indeed,
each target set contained three words with the
same valence, whereas Levens and Phelps (2008)
used emotional target sets of mixed neutral and
emotional stimuli. The memory for the emotional
stimuli was thus not increased and therefore PI
not reduced as it was in Levens and Phelps’study.
In the latter case, the reduction in PI for
emotional stimuli could be due to the distinctive-
ness of emotional stimuli among neutral stimuli,
this distinctiveness increasing memory and there-
fore reducing PI (assuming that when the source
memory for a stimulus is increased, the likelihood
for PI to occur is decreased, as hypothesised by
Levens and Phelps). Second, the time of stimuli
presentation was shortened, and third, the number
of stimuli was reduced. These modifications may
have decreased the source signal strength (less
time being dedicated to the encoding) and
improved the familiarity signal (the stimuli being
repeated more frequently). Thus, in contrast to the
findings by Levens and Phelps, our RNT may
have increased the difficulty in resolving PI for
emotional stimuli (at least for the positive).
In other words, modifications in the way emo-
tional stimuli stand out from their surroundings
appear to have led to differences in the way PI was
resolved for these stimuli. Indeed, special mechan-
isms of interference resolution with emotional stim-
uli may be expressed differently, in accordance with
the previous integration of salient context informa-
tion (i.e., all aspects that establish the overall
perceptual context; Attar, Müller, Andersen, Büchel,
&Rose,2010). Thus, although our results do not
run counter to the hypotheses of Levens and Phelps
(2008), regarding the basis of the emotional facilita-
tion of interference resolution, they convey a nuance
concerning the context in which they operate.
HOW EMOTION INTERFERES WITH INHIBITION
COGNITION AND EMOTION, 2014 7
Downloaded by [UCL Service Central des Bibliothèques] at 01:08 05 June 2014
We did not, however, observe the same effect for
RTs as for errors. Given the fact that resistance to PI
in the RNT requires the execution of several
mechanisms (e.g., competition among internal repre-
sentations, monitoring processes and response selec-
tion; Jonides & Nee, 2006), a possible explanation of
this dissociation is that emotion only dealt with
specific mechanisms of PI resolution such as
response selection (therefore, having an impact on
accuracy but not on response time).
Future studies are required to confirm that the
divergence between our results and the results
found by Levens and Phelps (2008) is actually due
to differences in paradigm. One direction for
upcoming experiments would be to use our RNT
design and that of Levens and Phelps, as well as
other experimental conditions manipulating the
time of stimuli presentation and the number of
stimuli. Future studies are also needed to further
clarify why we observed a significant effect of
valence only for errors. In addition, because of the
rather low number of errors, we might question
the sensitivity of the RNT used to assess PI in
working memory and reflect on the use of more
sensitive recent probe tasks (Nee, Jonides, &
Berman, 2007).
The current study adds to a growing literature
linking emotions and inhibitory control. Indeed,
although previous studies found that emotionally
laden stimuli interfere with the ability to inhibit a
prepotent response inhibition (Herbert & Sütter-
lin, 2011; Schulz et al., 2007; Verbruggen & De
Houwer, 2007) but facilitate resistance to PI
(Levens & Phelps, 2008), our results emphasised
a deleterious effect of emotions on these two
inhibitory functions. In addition, the finding that
valence interferes differently in accordance with
the task (i.e., greater interference of angry than joy
faces in the SST; interference of positive words
only in the RNT) supports the idea that distinct
mechanisms underlie these two inhibition-related
functions (i.e., top-down vs. bottom-up processes)
and possible consecutive differences in emotional
processing. Most of all, the current study high-
lights different interactions with stimulus/task
properties (e.g., pictures vs. words), processing
stages (e.g., early vs. later time course of emotional
processing) or motivational aspects (e.g., protec-
tion of the self vs. achievement or maintenance of
emotional well-being), yielding to possible differ-
ences in emotional processing. These interactions
further support the dual competition model (Pes-
soa, 2009), which proposes that emotion and
motivation affect both perceptual and executive
completion, leading to different behavioural per-
formances in accordance with the situation. This is
especially interesting for the comprehension of
self-regulation failures in which affect regulation
and impulse control conflict (Tice, Bratslavsky, &
Baumeister, 2001).
Manuscript received 22 October 2012
Revised manuscript received 21 February 2014
Manuscript accepted 4 May 2014
First published online 2 June 2014
REFERENCES
Attar, C. H., Müller, M. M., Andersen, S. K.,
Büchel, C., & Rose, M. (2010). Emotional pro-
cessing in a salient motion context: Integration of
motion and emotion in both V5/hMT+ and the
amygdale. Journal of Neuroscience,30, 5024–5210.
doi:10.1523/JNEUROSCI.5029-09.2010
Baumeister, R. F., Bratslavsky, E., Finkenauer, C., &
Vohs, K. D. (2001). Bad is stronger than good. Review
of General Psychology,5, 323–370. doi:10.1037/1089-
2680.5.4.323
Billieux, J., Gay, P., Rochat, L., & Van der Linden, M.
(2010). The role of urgency and its underlying
psychological mechanisms in problematic behaviours.
Behaviour Research and Therapy,48, 1085–1096.
doi:10.1016/j.brat.2010.07.008
Carstensen, L. L., & Mikels, J. A. (2005). At the
intersection of emotion and cognition: Aging and
the positivity effect. Current Directions in Psycho-
logical Science,14, 117–121. doi:10.1111/j.0963-
7214.2005.00348.x
Friedman, N. P., & Miyake, A. (2004). The relations
among inhibition and interference control functions: A
latent-variable analysis. Journal of Experimental Psycho-
logy: General,133, 101–135. doi:10.1037/0096-3445.
133.1.101
Gay, P., Rochat, L., Billieux, J., d’Acremont, M., &
Van der Linden, M. (2008). Heterogeneous inhibi-
tion processes involved in different facets of self-
REBETEZ ET AL.
8COGNITION AND EMOTION, 2014
Downloaded by [UCL Service Central des Bibliothèques] at 01:08 05 June 2014
reported impulsivity: Evidence from a community
sample. Acta Psychologica,129, 332–339. doi:10.1016/j.
actpsy.2008.08.010
Herbert, C., & Sütterlin, S. (2011). Response inhibi-
tion and memory retrieval of emotional target
words: Evidence from an emotional stop-signal
task. JournalofBehavioralandBrainScience,1,
153–159. doi:10.4236/jbbs.2011.13020
Jonides, J., & Nee, D. E. (2006). Brain mechanisms of
proactive interference in working memory. Neuroscience,
139, 181–193. doi:10.1016/j.neuroscience.2005.06.042
Levens, S. M., & Phelps, E. A. (2008). Emotion
processing effects on interference resolution in work-
ing memory. Emotion,8, 267–280. doi:10.1037/
1528-3542.8.2.267
Logan, G. D. (1994). On the ability to inhibit thought
and action: A user’s guide to the stop signal
paradigm. In D. Dagenbach, & T. H. Carr (Eds.),
Inhibitory processes in attention, memory, and language
(pp. 189–239). San Diego, CA: Academic Press.
Lundqvist, D., Flykt, A., & Öhman, A. (1998). The
Karolinska directed emotional faces. Stockholm: Kar-
olinska Institute, Psychology Section, Department of
Clinical Neuroscience.
Nee, D. E., Jonides, J., & Berman, M. G. (2007).
Neural mechanisms of proactive interference-res-
olution. NeuroImage,38,740–751. doi:10.1016/j.
neuroimage.2007.07.066
Öhman, A., & Mineka, S. (2001). Fears, phobias, and
preparedness: Toward an evolved module of fear and
fear learning. Psychological Review,108, 483–522.
doi:10.1037/0033-295X.108.3.483
Pessoa, L. (2009). How do emotion and motivation
direct executive control? Trends in Cognitive Sciences,
13, 160–166. doi:10.1016/j.tics.2009.01.006
Phillips,M.L.,Drevets,W.C.,Rauch,S.L.,&LaneR.
(2003). Neurobiology of emotion perception II:
Implications for major psychiatric disorders. Biological
Psychiatry,54, 515–528. doi:10.1016/S0006-3223(03)
00171-9
Schimmack, U. (2005). Attentional interference effects
of emotional pictures: Threat, negativity, or arousal?
Emotion,5,55–66. doi:10.1037/1528-3542.5.1.55
Schulz, K. P., Fan, J., Magidina, O., Marks, D. J.,
Hahn, B., & Halperin, J. M. (2007). Does the
emotional go/no-go task really measure behavioral
inhibition? Convergence with measures on a non-
emotional analog. Archives of Clinical Neuropsychol-
ogy,22, 151–160. doi:10.1016/j.acn.2006.12.001
Schupp, H. T., Flaisch, T., Stockburger, J., &
Junghöfer, M. (2006). Emotion and attention:
Event-related brain potential studies. Progress in
Brain Research,156,31–51. doi:10.1016/S0079-
6123(06)56002-9
Talmi, D., Luk, T. C. B., McGarry, L., & Moscovitch,
M. (2007). The contribution of relatedness and
distinctiveness to emotionally-enhanced memory.
Journal of Memory and Language,56, 555–574.
doi:10.1016/j.jml.2007.01.002
Tice, D. M., Bratslavsky, E., & Baumeister, R. F.
(2001). Emotional distress regulation takes preced-
ence over impulse control: If you feel bad, do it!
Journal of Personality and Social Psychology,80,53–67.
doi:10.1037/0022-3514.80.1.53
Verbruggen, F., & De Houwer, J. (2007). Do emotional
stimuli interfere with response inhibition? Evidence
from the stop signal paradigm. Cognition and Emo-
tion,21, 391–403. doi:10.1080/02699930600625081
Yoon, K. L., Joorman, J., & Gotlib, I. H. (2009).
Judging the intensity of facial expressions of
emotion: Depression-related biases in the proces-
sing of positive affect. Journal of Abnormal Psycho-
logy,118, 223–228. doi:10.1037/a0014658
HOW EMOTION INTERFERES WITH INHIBITION
COGNITION AND EMOTION, 2014 9
Downloaded by [UCL Service Central des Bibliothèques] at 01:08 05 June 2014
