ACTPSY-01481; No of Pages 15
ARTICLE IN PRESS
March 25, 2010;
Model: Guliver 5
Acta Psychologica xxx (2010) xxx–xxx
Contents lists available at ScienceDirect
Acta Psychologica
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / a c t p s y
Emotional valence of spoken words influences the spatial orienting of attention☆
Julie Bertels a,b,⁎, Régine Kolinsky a,b, José Morais c
a
b
c
Université Libre de Bruxelles (U.L.B.), Belgium
Fonds de la Recherche Scientifique – FNRS (F.R.S.–FNRS), Belgium
Université Libre de Bruxelles (U.L.B.), Belgium
a r t i c l e
i n f o
Article history:
Received 18 December 2008
Received in revised form 23 February 2010
Accepted 24 February 2010
Available online xxxx
PsycINFO classification:
2346 Attention
2360 Motivation & Emotion
Keywords:
Auditory attention
Cognitive bias
Emotional content
Spoken words
a b s t r a c t
The influence of the affective content of speech on the spatial orienting of auditory attention was examined
by adapting the dot probe task. Two words, one of which was emotional in one quarter of the trials, were
played simultaneously from a left- and a right-located loudspeaker, respectively, and followed (or not) by a
lateralized beep to be detected (Experiments 1 and 2) or localized (Experiment 3). Taboo words induced
attentional biases towards their spatial location in all experiments, as did negative words in Experiment 3,
but not positive words. Thus, in audition, the identification of an emotional word automatically activates the
information about its spatial origin. Moreover, for both word types, attentional biases were only observed
when the emotional word was presented on the participant's right side, suggesting that the dominant left
hemisphere processing of words constrains the occurrence of spatial congruency effects.
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
The present work examines whether the affective content of speech
influences the spatial orienting of auditory attention. If so, an emotional
word might facilitate or interfere with the processing of an unrelated,
simple auditory signal subsequently presented, depending on whether
this signal shares the spatial origin of the speech stimulus or not.
It has been proposed that the auditory system, like the visual one,
is composed of two parallel streams of processing: the “what” and the
“where” subsystems (Kubovy & Van Valkenburg, 2001). The dissociation of these two subsystems is supported, for example, by the
observation of an illusion in which a sequence of pitches is perceived
correctly but in the wrong ear location (Deutsch & Roll, 1976), and by
the fact that adaptation may have a detrimental effect on localization
while leaving pitch perception unaltered (Hafter, 1997). The two
auditory subsystems seem to rely on distinct cerebral structures: the
“what” subsystem relying on the superior temporal cortex and the
☆ The first author was a Research Fellow of the Fonds de la Recherche Scientifique
(F.R.S.–FNRS) and is presently a Scientific Research Worker of the F.R.S.–FNRS. The second
author is a Senior Research Associate of the F.R.S.–FNRS. All authors were affiliated to the
Research Unit in Cognitive Neurosciences, Université Libre de Bruxelles.
⁎ Corresponding author. UNESCOG, Université Libre de Bruxelles, Avenue F.D.
Roosevelt, 50, C.P.191, 1050 Brussels, Belgium.
E-mail address: [email protected] (J. Bertels).
“where” subsystem on the medial, posterior auditory cortex (Clarke,
Bellmann, Meuli, Assal, & Steck, 2000; Rauschecker, 1997, 1998a,b).
In audition, as perceptual form and organization are not processed
in relation to space, possible links between the “what” and the
“where” auditory subsystems have been underestimated (see Kubovy,
1988). Still, it can be argued that such functional links afford us an
ecological advantage. In peripheral space, beyond a 20° visual angle,
auditory targets elicit shorter orientation latencies than visual targets
(Goldring, Dorris, Corneil, Ballantyne, & Munoz, 1996) and, at night,
the identification of sounds signalling either danger or urgent
requests for help may require auditory localization. Consequently,
identification of an auditory object relevant for survival (or at least for
ecological adaptation) might automatically activate information
about the spatial location of its source in order to orient spatial
attention to the same location. This would put the listener in a better
attentional state to process subsequent auditory (or visual) information coming from the same source. For example, identifying the kind
of animal that produced a particular cry should, in principle, influence
the level of attention devoted to its position in space and therefore the
accuracy and/or rapidity of detection or localization of subsequent
signals coming from the same direction.
Obviously, not all auditory objects convey shallow, non-symbolic,
information. Spoken words are auditory objects that transmit highly
abstract information, including semantics. As regards spoken words,
some refer to emotional states or are emotionally loaded. Although
emotional words are not more directly relevant for survival than
0001-6918/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.actpsy.2010.02.008
Please cite this article as: Bertels, J., et al., Emotional valence of spoken words influences the spatial orienting of attention, Acta Psychologica
(2010), doi:10.1016/j.actpsy.2010.02.008
ARTICLE IN PRESS
2
J. Bertels et al. / Acta Psychologica xxx (2010) xxx–xxx
neutral words (especially when uttered in a neutral intonation), their
identification might evoke emotional feelings, boost the attentional
system and trigger the “where” auditory subsystem, influencing the
orientation of spatial attention.
Many authors examined the spatial orienting of auditory attention,
namely the characteristics of the auditory “where” subsystem,
without considering the involvement of the “what” subsystem (e.g.,
Buchtel & Butter, 1988; Rhodes, 1987; Spence & Driver, 1994). Most of
these studies used an auditory variant of the cuing paradigm (Posner,
1980), in which a peripherally-presented auditory cue is followed by
an auditory target, presented either at the same location as the cue or
at the opposite location (in valid and invalid trials, respectively).
Exogenous attentional orienting to the spatial location of the cue is
reflected by the validity effect, namely the reaction time (henceforth,
RT) difference between invalid and valid trials.
In the present study, we used a variant of the auditory cuing
paradigm to attempt to assess the impact of the emotional content of
word cues, relying on the “what” subsystem, on the spatial orienting
of auditory attention.
This variant actually consists of an adaptation to the auditory
modality of the dot probe task (also called attentional deployment
task, MacLeod, Mathews, & Tata, 1986) used in the visual modality
to examine the role of the emotional content of written words on
the spatial orienting of visual attention. In the dot probe task, two
written words, either one emotionally loaded and the other not, or
both neutral, are presented one above the other. In half of the
trials, the words are followed by an unrelated visual target (a dot)
in the location that was previously occupied by one of the words.
On congruent trials (i.e. valid trials, in the terminology of the cuing
paradigm), the dot is presented at the same location as the one
previously occupied by the emotional word, while on incongruent
(i.e. invalid) trials the dot is presented at the opposite location. In
our auditory adaptation of the dot probe task (named the beep
probe task), during each trial two different spoken words were
presented through loudspeakers in a diotic way, i.e. simultaneously, one on the left and the other on the right side of the
listener. The use of loudspeakers instead of headphones was
motivated by the ecological advantage of making sounds come
from the external world. In the relevant trials, one of the words
was emotional and the other was neutral. In half of the trials, a
lateralized single beep target followed the presentation of the pair
of words. The use of an auditory target instead of a visual one, as
in the dot probe task, allowed us to keep the unimodal aspect of
the original paradigm.
The congruence or incongruence between the presentation side
of the emotional word and the presentation side of the beep was
critical. Consistent with the underlying idea of the cuing paradigm
that the faster the participant reacts to a given location, the most
attention was attracted in this location (Posner, Snyder, &
Davidson, 1980), the prediction derived from the auditory “what–
where” connection hypothesis was that the spatial congruence
between the emotional word and the beep would lead to shorter
RTs than their spatial incongruence, hence creating what is called
an attentional bias, following MacLeod and Mathews' (1988)
terminology.
It must be noted that the auditory “what–where” connection
hypothesized here needs not be emotion-dependent. However, as in
the present situation only one of the diotic stimuli was emotional, any
spatial congruency effect implies that the “what–where” connection is
driven (or at least is reinforced) by the emotional member of the
diotic pair, not by the neutral one.
Given that both the words and beeps were presented in a
lateralized way, perceptual laterality effects related to hemispheric
specialization were likely to be observed. Indeed, as shown by
Demakis and Harrison (1994), both neutral and affective words are
mainly or more efficiently processed by the left hemisphere, at least
when the emotional content of the word is irrelevant to the task.
Furthermore, as Kinsbourne (1970) argued, the presentation of verbal
material could activate the left hemisphere preponderantly, thus
leading to involuntary orientational biases to the right side of space.
To this hemispheric imbalance hypothesis one may add the
assumption that an emotional word catches more attentional
resources than its paired neutral word, and therefore activates the
hemispheric contralateral to its side of presentation to a greater
extent than the ipsilateral one. Thus, taking these two assumptions
together, one may predict the attentional bias to be stronger when the
emotional word is presented on the right side. Indeed, in this case, the
processing of a right-sided beep would be favoured by both
attentional mechanisms, the “hemispheric” and the “emotional”
one, whereas processing of a left-side beep would be disadvantaged
by both. In contrast, if the emotional word is presented on the left side,
processing a beep presented on either the left or the right side would
benefit from one of the two mechanisms, either the “emotional” or the
“hemispheric” one, respectively, but would be disadvantaged by the
other mechanism.
The influence of the emotional content of a word on attentional
orienting could depend on its specific emotional valence. Because the
rapid processing of positive information does not seem to be as crucial
for the organism's well-being as is the processing of negative
information (Baumeister, Bratslavsky, Finkenauer, & Vohs, 2001;
Grühn, Smith & Baltes, 2005; Rozin & Royzman, 2001; Taylor, 1991),
we hypothesized the “where” auditory subsystem to be more likely
triggered by negative and taboo words (namely vulgar, shocking or
obscene words, as well as insults) than by positive ones. Hence, we
expected taboo and negative spoken words to facilitate the processing
of the beep presented at the same spatial location (compared to its
presentation at the opposite side) to a greater extent than positive
words. This effect would be larger for right-presented emotional
words, as argued above.
Experiment 1 consisted of an auditory adaptation of the original
version of the dot probe task. As in this task, the participants were
required to detect the presence of the non-linguistic target in each
trial, here the beep. Also, given that in the visual paradigm
participants were asked to read the word presented in the upper
position in order to be sure they paid attention to the linguistic
stimuli, we asked them to repeat the word presented through a prespecified loudspeaker after responding to the beep in Experiment 1.
Given that only effects of taboo words emerged, we eliminated this
repetition task in a second experiment, speculating that it might have
monopolized some attentional resources, making it difficult for
negative words to facilitate selectively the detection of a same-side
beep. In addition, the attentional biases observed in Experiment 1 for
taboo words could have been due to the fact that participants had to
pronounce these shocking words in front of the experimenter.
Elimination of the repetition task thus allowed us to both reduce
the cognitive load and remove this potential bias. Finally, in a third
experiment, we used a localization rather than a detection task in
order to introduce a spatial component in the task, given that spatial
validity effects have been observed more frequently in the auditory
cuing paradigm when spatial representations are somehow relevant
for the participant's goal (Rhodes, 1987; Schmitt, Postma, & De Haan,
2000; Spence & Driver, 1994).
2. Experiment 1: Beep probe detection and word repetition
This experiment was intended to be a close auditory analogue of
the classic version of the dot probe task (MacLeod et al., 1986, see also
e.g., Fox, 1993 and Salemink, van den Hout, & Kindt, 2007), using
lateral stimulations. We did not use up–down stimulations because
these proved to be exceedingly hard to discriminate. Neutral,
negative, positive and taboo words were presented. Participants
were asked to detect the target — here, a beep — and to then repeat a
Please cite this article as: Bertels, J., et al., Emotional valence of spoken words influences the spatial orienting of attention, Acta Psychologica
(2010), doi:10.1016/j.actpsy.2010.02.008
ARTICLE IN PRESS
J. Bertels et al. / Acta Psychologica xxx (2010) xxx–xxx
word of the pair presented on a pre-specified side, in the same way
that they had to read aloud the upper word of the pair in the original
version of the dot probe task in order to make sure they paid attention
to the linguistic stimuli.
Following the auditory cuing literature, the use of a detection
task would prevent the occurrence of any validity effect (Rhodes,
1987; Schmitt et al., 2000), at least when loudspeakers and
uninformative cues are used (Spence & Driver, 1994). Indeed, the
lack of any spatial representation linked to the task would probably
preclude spatial orienting towards the location of the auditory cue.
Nevertheless, Buchtel, Butter, and Ayvasik (1996) argued that
validity effects are observed with a detection task when there is
an orienting reaction towards the source of the sound, namely in
conditions favouring the intention to orient the head to a sound
source. Following this idea, given the instruction to repeat a word
presented on a pre-specified loudspeaker on each trial (i.e. to orient
attention voluntarily towards one side of the auditory space) and in
spite of the fact that a detection task and uninformative cues were
used in Experiment 1, we expected to observe congruency effects
between the position of the emotional word in the pair and the
position of the following beep.
As any activation of the “what–where” connections is presumably
due to the need to register relatively rapid changes in the
environment, we assumed that it would be transient. To test this
idea we used two word-beep intervals here, one short and one long,
and expected the predicted attentional bias (i.e. facilitated processing
of the beep presented at the same spatial location as the emotional
word of the pair, compared to the beep presented at the opposite side)
to occur with the short interval, not — or at least less so — with the
long one. Moreover, if an attentional bias were to occur in the long
interval condition, we predicted it to be negative. Namely, participants would detect the beep faster if presented at the opposite
location (i.e. at the location of the neutral word) than if it was
presented at the same location as the emotional word. As a matter of
fact, given the lengthening of the SOA (Stimulus Onset Asynchrony,
i.e. the time between the onset of the word and the onset of the beep)
in this condition, one could expect an inhibition of return phenomenon (IOR, cf. Posner & Cohen, 1984) to occur: with long SOAs,
attention oriented to the location of the cue would come back to a
central location for the application of inhibitory processes to the
location previously explored, making it easier to process any
information coming at the opposite rather than at the same location
as the cue.
2.1. Method
2.1.1. Participants
Participants were 35 first-year students of the Université Libre
de Bruxelles (31 women; 3 left-handed), ranging from 17 to
40 years (mean: 20.3). They were given course credits for their
participation. The results of three women, all right-handed, were
discarded from the analyses: two because their average RTs (all
experimental conditions mingled) were superior to two standard
deviations above the overall average performance, and one because
of her inability to repeat correctly the presented words. All
participants had spoken French for at least the last 10 years.
Previous studies investigating attentional biases in the visual
modality pointed to the robustness of this phenomenon only in
selected groups of participants such as anxious subjects (e.g., BarHaim, Lamy, Pergamin, Bakermans–Kranenburg, & van Ijzendoorn,
2007), repressors (Mogg et al., 2000) and depressed people (e.g.,
Mogg, Bradley & Williams, 1995). Therefore, we used personality
questionnaires (filled in after the beep probe task) evaluating
anxiety and depression levels (the Spielberger Trait-State Anxiety
Inventory, STAI-Y; Spielberger, 1983, and the Beck Depression
Inventory, BDI–II; Beck, Steer, & Brown, 1996, respectively), as well
3
as willingness to be socially desirable (the Marlowe–Crowne Scale
of Social Desirability, SDS; Crowne & Marlowe, 1960) in order to
examine whether these measures would correlate with the
emotional effects. Table 1 displays the average scores obtained in
these questionnaires.
Participants were randomly assigned to one of four groups (8 participants in each group) created on the basis of the pair-beep interval
length (short/long) and the side of the auditory space (the
loudspeaker) to which they had to pay attention in order to repeat
the word presented on that side (left/right), hereafter called repetition
side.
2.1.2. Material and apparatus
Words were orally presented. They were spoken by a Frenchspeaking female theatre student in a neutral tone of voice and were
digitally recorded on a Sony MiniDisc. Stimuli were then transferred
on a Macintosh Powerbook G3 via the Digidesign DIGI 002 Rack
interface and were cleaned and normalized with the Protools
Digidesign 6.2.2. software. Mean word duration was 719 ms (standard
deviation: 135 ms).
The words used in this experiment (as in the following ones)
were chosen as described below. There were 480 pairs of mono- or
bi-syllabic words: 120 emotional pairs, in which one of the words
was emotionally neutral and the other was emotionally charged (see
Appendix A), and 360 neutral pairs, consisting of two emotionally
neutral words. Emotional pairs included 40 positive (e.g., amour/
agence — love/agency), 40 negative (e.g., bombe/bulle — bomb/
bubble) and 40 taboo pairs (e.g., salope/serrure — bitch/lock), defined
as such because the emotional word in the pair was positive,
negative or taboo. These words were generated by collaborators or
chosen from the relevant literature (Bonin et al., 2003; Hermans &
De Houwer, 1994; Messina, Morais & Cantraine, 1989; Van der
Goten, De Vooght & Kemps, 1999). In accordance with the original
version of the dot probe task (e.g. MacLeod et al., 1986, see also Fox,
1993 and Salemink et al., 2007), all the 120 emotional pairs as well
as 120 neutral pairs (i.e., as many neutral as emotional pairs) were
critical pairs, as they were followed by a beep. The others were filler
pairs, also considered as catch trials, since there was no beep to
detect.
On the basis of the database Lexique 2 (New, Pallier, Ferrand, &
Matos, 2001), each category of emotional words was matched to
the set of neutral words associated with them in a pair, according
to the number of phonological neighbours, literary frequency and
web frequency. The use of web frequency was crucial for taboo
words given that it better reflects their frequency of use than the
literary frequency. Hence, these matchings ensure that attentional
orienting towards the spatial location of either type of word is not
due to these lexical factors. According to the same criteria, we also
matched all emotional words to all associated neutral words, and
the two series of neutral words constituting the neutral critical
pairs. Finally, we matched these four sets of material. These
matching sets led to non-significant difference at 0.05. Within a
Table 1
Average standard scores obtained on the STAI-Y (Spielberger, 1983) and on the SDS
(Crowne & Marlowe, 1960) and average raw scores obtained on the BDI–II (Beck et al.,
1996) in Experiments 1, 2 and 3 (standard deviations are in parentheses).
STAI-Y
State anxiety
Trait anxiety
SDS
BDI-II
Exp. 1 (n = 32)
Exp. 2 (n = 24)
Exp. 3 (n = 20)
48.09
48.84
47.72
10.56
46.38
49.58
47.83
12.08
48.05
46.75
46.96
9.4
(10.80)
(10.36)
(22.35)
(7.8)
(7.46)
(10.81)
(12.01)
(8.54)
(6.79)
(9.54)
(7.64)
(5.59)
Please cite this article as: Bertels, J., et al., Emotional valence of spoken words influences the spatial orienting of attention, Acta Psychologica
(2010), doi:10.1016/j.actpsy.2010.02.008
ARTICLE IN PRESS
4
J. Bertels et al. / Acta Psychologica xxx (2010) xxx–xxx
pair, the first phoneme, the syllabic structure and the phonological
uniqueness point of the two words were identical. Words of the
filler pairs were only matched in terms of first phoneme and
syllabic structure.
Four post-hoc control studies were carried out on these words or
part of them.
In the first post-hoc study, the 480 words constituting the critical
pairs, a priori classified as neutral (360), positive (40), negative (40)
and taboo (40), were presented to 24 listeners, who were asked to
rate each word on a 1 (very negative) to 7 (very positive) scale for
emotionality. The difference between the emotional means of the two
words of each pair was calculated from these ratings. For emotional
pairs of words the minimal acceptable difference was one point,
whereas for neutral pairs the difference had to be smaller than one
point. Following this criterion, two negative, seven positive and 14
neutral pairs were removed from analyses.
In the second post-hoc study, a list of 120 words a priori
classified as neutral (80) and taboo (40) was presented to 24
listeners, who were asked to rate each word on a 1 (normal) to 5
(very taboo, shocking) scale for obscenity. No word was removed
from analyses on the basis of these ratings, as the confidence
intervals of the two word categories did not overlap and as the
minimum and the maximum values corresponding to one category
did not enter into the confidence interval of the other category;
however, one taboo word was removed because more than the half
of the participants did not understand its shocking meaning
(biroute).
The characteristics of the emotional pairs that were taken into
account in the statistical analyses are presented in Table 2. Since the
shock value was only assessed for the taboo words and part of the
neutral words, only data concerning the shock value of the taboo
words are mentioned in Table 1. It should be noted that the exclusion
of 24 pairs on the basis of their emotional valence or the
understanding of their meaning (see above) did not affect the
matching of the three categories of emotional pairs, on literary
frequency, web frequency or the number of phonological neighbours.
Results from the two subsequent control studies were indicative
and did not serve to exclude pairs from further analyses. In the
third post-hoc study, 24 participants were asked to rate the 240
words from the emotional pairs, a priori classified as neutral (120),
negative (40), positive (40) and taboo (40), on a 1 (very quiet) to 7
(very arousing) arousal scale. Regarding words presented in the
non-excluded pairs, statistical analysis showed that the arousal
level differed between the emotional types of words, F(2, 109) =
450.29, p b 0.001, η2 = 0.894, each type differing from the others,
p b 0.001 for all (Bonferroni corrected), with negative words
producing the highest arousal, followed by the taboo and the
positive words. For each pair we also calculated the difference in
arousal level between the emotional word and its neutral associate.
The effect of the type of emotional pair was significant, F(2, 109) =
258.2, p b 0.001, η2 = 0.828, each type differing from all the others
at p b 0.001 (Bonferroni corrected). Negative pairs led to the
greatest difference (mean: 2.04), followed by taboo (1.37) and
positive pairs (−1.64).1 It is also worth noting that the shock value
1
As we developed in a normative study (Bertels, Kolinsky, & Morais, 2009), it is not
surprising that positive words were judged as less arousing than their associate
neutral words given that in the 7-point scale we used the “1” level depicting a very
calm, relaxing feeling; the “7” level reflecting a very arousing, exciting feeling; and the
“4” level referring to no particular feeling. This actually means that while neutral
words were judged (on average) as not evoking any particular arousing feeling,
positive words were in fact judged as more calming and relaxing than neutral words
(see Table 2).
of the taboo words was positively correlated with their arousal
level, r = 0.56, p b 0.001.
In the fourth post-hoc study, 24 participants rated the same
words as in study 3 on a 1 (unfamiliar) to 5 (very familiar) scale for
familiarity. Regarding words presented in the non-excluded pairs,
statistical analyses revealed that familiarity differed between the
emotional types of words, F(2, 109) = 66.679, p b 0.001, η2 = 0.555,
with positive words more familiar than the negative and taboo
words, which also differed from each other (all p ≤ 0.001,
Bonferroni corrected). We also calculated for each pair the
difference in familiarity between the emotional word and its
neutral associate. The effect of type of emotional pair was
significant, F(2, 109) = 3.36, p = 0.04, η2 = 0.059, reflecting that
positive pairs led to a greater difference than negative (0.45 vs.
0.08, p = 0.04, Bonferroni corrected), but not taboo pairs (0.34,
F b 1). The difference between negative and taboo pairs did not
reach significance (p N 0.10).
In the beep probe task, each pair of words was presented
diotically, namely with the two words constituting the pair
presented simultaneously with each different member spoken
through a different, lateralized, loudspeaker. Words within each
pair were synchronized at their onset and offset, which was
facilitated by the fact that they had the same syllabic structure, and
were uttered at a constant rate. This synchronisation was made by
operating very short excisions using the Protools Digidesign 6.2.2.,
and was perceptually evaluated by the first author. Critical pairs
were followed by a 100-ms beep. Filler pairs were followed by a
silence of identical duration.
2.1.3. Procedure
Participants sat in front of a computer. Two loudspeakers were
located at 45 cm on their left and right, with an anterior deviation
of 60° in relation to the sagittal plane. Stimulus presentation and
timing as well as data collection were controlled using the
Psyscope button box and 1.2.5. PPC software (Cohen, MacWhinney,
Flatt, & Provost, 1993) running on a Macintosh Performa 6320.
Each session began with detailed instructions. Participants were
told that they would hear pairs of different words, presented
simultaneously, one on the left and one on the right, and that in
half of the trials, a beep would be presented after the pair of words,
either on the left or the right. They had to perform two tasks on
each trial: firstly, to detect the beep by pressing the right key of a
button box as quickly and accurately as possible with the preferred
hand, without caring about its side of presentation; and secondly,
to repeat as accurately as possible the word presented at a prespecified side (the same throughout the testing), to ensure they
paid attention to the presented words.
Each trial started with the presentation of a 750 ms fixation
cross in the middle of the screen. Then, a pair of auditory words
was presented. In half of the trials, immediately or 250 ms after the
offset of the words (depending on the participants), a beep was
presented, at the location of one of the words. The participant had
2500 ms to answer. The interval between the response and the
next trial was 2000 ms. Trials were pseudo-randomly presented: a
beep was presented in not more than three trials in a row, and
similarly a word of the same emotional type was presented in not
more than three trials in a row whatever the presentation side, and
never more than in two trials in a row on the same presentation
side.
The experiment started with a 36-trial practice block, during
which participants received feedback regarding their performance.
Next, each participant was presented with five 96-trial blocks,
without any feedback. Thus, a participant was presented with each
pair only once insofar as assignment of a specific word to one side
(e.g., “beauté–bateau” vs. “bateau–beauté”) and beep location were
counterbalanced between participants.
Please cite this article as: Bertels, J., et al., Emotional valence of spoken words influences the spatial orienting of attention, Acta Psychologica
(2010), doi:10.1016/j.actpsy.2010.02.008
ARTICLE IN PRESS
J. Bertels et al. / Acta Psychologica xxx (2010) xxx–xxx
5
Table 2
Characteristics of the emotional pairs taken into account in the statistical analyses, following post-hoc control studies.
n
Literary frequency
Negative pairs
Negative words
Neutral words
Positive pairs
Positive words
Neutral words
Taboo pairs
Taboo words
Neutral words
38
32.163
30.046
33
82.42
248.035
39
5.676
5.195
Web frequency
Number of phonological neighbours
Emotional valence
Shock value
Arousal
Familiarity
4261.413
9021.363
7.053
6.658
2.08
4.445
–
–
5.81
3.774
3.83
3.751
15,059.057
16,487.881
10.061
11.364
6.16
4.433
–
–
2.191
3.828
4.467
4.013
363.245
1385.998
8.769
7.41
2.877
4.386
2.984
–
5.052
3.686
3.549
3.209
2.2. Results
Errors (misses and false alarms) were rare, 0.28% on average.
Therefore, the analyses focused on RT data only. RTs on erroneous
trials as well as on trials in which the participant had not correctly
repeated the word (in total, 11.18% of the observations) were
excluded from further analyses. This was also the case for RTs on
the 24 pairs for which (a) the emotional word did not differ
considerably on emotional valence from the neutral word; or
(b) one neutral word did differ on emotional valence from the
other neutral word of the pair; or (c) the shocking meaning was
not straightforward (see Method).2 Table 3 displays the mean
correct RTs.
Firstly, we examined whether the presence of an emotional word
of a particular valence affected the response to the beep. For this
purpose, a 2 (interval length: short/long) × 2 (repetition side: left/
right) × 4 (type of emotional pair: neutral/negative/positive/taboo)
× 2 (beep location: left/right) repeated measures analysis of variance
(ANOVA) design was applied on correct response latencies. The type
of emotional pair and the beep location were within-subject factors,
interval length and repetition side were between-subject factors.
The only significant factor was the type of emotional pair, F(3, 84) =
17.8, p b 0.01, partial η2 = 0.389: taboo pairs led to significantly longer
RTs (463 ms, on the average) than neutral (416 ms), positive (415 ms)
and negative (414 ms) pairs respectively, all p b 0.01 (Bonferroni
corrected). No other comparison was significant, all p N 0.10. The
interaction between type of emotional pair and interval length did not
reach significance, F b 1.
The second important point was to assess whether emotional
words led to attentional biases, namely if they facilitated the detection
of the beep when this was presented at the same spatial location (in
spatially congruent trials) as the emotional word in comparison to the
situation where it was presented at the opposite spatial location (in
spatially incongruent trials). As in other studies that looked for
attentional biases linked to the emotional valence of the stimuli, the
analyses only took the results on emotional pairs into account, given
that, by definition, emotional word location is not available for neutral
pairs. A 2 (interval length: short/long) × 2 (repetition side: left/
right) × 3 (type of emotional pair: positive/negative/taboo) × 2 (beep
location: left/right) × 2 (emotional word location: left/right) repeated
measures ANOVA design was applied on response latencies. Interval
length and repetition side were between-subject factors, the other
variables were within-subject factors. The only main effect that
reached significance was the type of emotional pair, F(2, 56) = 18.46,
p b 0.001, partial η2 = 0.397. As with the previous analysis, Bonferroni
2
Note that results from the statistical analyses on the whole set of data were similar
to the ones obtained after exclusion of these pairs. In particular, the four-way
interaction between type of emotional pair, beep location, emotional word location
and interval length was significant, F(2, 56) = 4.37, p b 0.02, partial η2 = 0.135.
adjusted post-hoc comparisons showed that taboo pairs led to longer
RTs than negative and positive pairs, both p b 0.001, which did not
differ between each other, p N 0.10.
More crucially, there was a significant three-way interaction
between type of emotional pair, beep location and emotional word
location, F(2, 56) = 5.07, p b 0.01, partial η2 = 0.153, superimposed by
a significant four-way interaction between type of emotional pair,
beep location, emotional word location and interval length, F(2, 56) =
3.27, p b 0.05, partial η2 = 0.104. Considering each interval condition
separately, we found the effect of type of emotional pair to be
significant in both, F(2, 28) = 14.27 and 6.53, both p b 0.01, partial
η2 = 0.505 and 0.318. In contrast, the three-way interaction between
type of emotional pair, beep location and emotional word location
was significant in the short interval condition, F(2, 28) = 7.97,
p = 0.002, partial η2 = 0.363, but not in the long one, F b 1.
In the short interval condition, we analysed separately the data for
each type of emotional pair, given that we predicted the occurrence of
attentional biases with taboo and negative pairs, but not with positive
pairs. The interaction between beep location and emotional word
location was significant for the taboo pairs, F(1, 14) = 9.46, p b 0.01,
partial η2 = 0.403, but not for the negative pairs, F(1, 14) = 1.8, p N 0.10,
partial η2 = 0.114; for positive pairs there was only a trend, F(1, 14) =
3.7, p = 0.08, partial η2 = 0.208. This interaction reflects the fact that,
when the taboo word was right-presented, participants detected the
beep faster when this was right-presented (434 ms) than when it was
left-presented (515 ms), p b 0.001 (Bonferroni corrected). No effect
was observed when the taboo word was left-presented, p N 0.10 (see
Fig. 1). Thus, the attentional bias reflected by the RT difference (81 ms)
Table 3
Mean correct detection times observed in Experiment 1 for all types of pairs (standard
errors are in parentheses).
Type of emotional pair
Beep and emotional word
locations
Short interval condition
Beep Left
Emotional word Left
Emotional word Right
Beep Right
Emotional word Left
Emotional word Right
Average
Long interval condition
Beep Left
Emotional word Left
Emotional word Right
Beep Right
Emotional word Left
Emotional word Right
Average
Neutral
Negative
Positive
Taboo
448 (28)
402 (17)
435 (21)
407 (19)
468 (27)
515 (29)
429 (23)
431 (24)
427 (18)
407 (22)
431 (29)
420 (18)
489 (33)
434 (27)
477 (24)
394 (16)
397 (20)
409 (17)
398 (20)
428 (18)
459 (33)
406 (25)
403 (22)
400 (18)
410 (23)
422 (33)
410 (18)
444 (28)
467 (37)
450 (24)
436 (21)
424 (20)
430 (17)
409 (17)
395 (21)
402 (17)
Please cite this article as: Bertels, J., et al., Emotional valence of spoken words influences the spatial orienting of attention, Acta Psychologica
(2010), doi:10.1016/j.actpsy.2010.02.008
ARTICLE IN PRESS
6
J. Bertels et al. / Acta Psychologica xxx (2010) xxx–xxx
Fig. 1. Target detection latencies observed for taboo pairs in Experiment 1, in the short
interval length condition, as a function of the emotional word and beep locations.
between the spatially incongruent and congruent taboo trials was
restricted to the condition in which the taboo word was rightpresented. This bias was not correlated with any of the scores on the
personality questionnaires (STAI-state: r = −0.15; STAI-trait: r = 0.11;
BDI–II: r = −0.05; SDS: r = 0.07, all p N 0.50).
Moreover, the attentional bias linked to right-presented taboo
words in the short interval condition was significantly larger than the
non-significant biases associated with right-presented negative and
positive words (−30 and −24 ms, respectively), t(15) = 3.753 and
4.123, both p b 0.005 (Bonferroni corrected).
It is of theoretical interest to consider on which spatial attentional
orienting component(s) taboo words had an effect. A positive attentional
bias like the one observed here for right-presented taboo words (i.e., a
positive difference between RTs on incongruent and congruent trials for
these words) could be due either to vigilance to the taboo words,
namely an effect on the engagement component of attentional
orienting, or to difficulties in disengaging attention from these words
once attention has been normally engaged. In the former case, the
attentional bias would result from a RT advantage on congruent
compared to incongruent trials, whereas in the latter case it would
proceed from increased RTs on incongruent trials. However, merely
comparing RTs on congruent and incongruent trials does not disentangle these possibilities given the absence of a baseline in the emotional
trials. To solve this problem, Koster, Crombez, Verschuere, and De
Houwer (2004) proposed to use the neutral trials (i.e., trials in which
both words are emotionally neutral) as baseline and to compare them
systematically with the RTs on congruent and incongruent trials, but this
approach has been recently criticized (Mogg, Holmes, Garner, & Bradley,
2008). Given that neutral trials do not, by definition, contain emotional
words, these authors argued that potential non-spatial interference
effects of emotional words could contribute to the RT difference
between neutral and spatially (in)congruent trials. As such general,
non-spatial effects of taboo words occurred in the present experiment,
the RT difference with neutral words would not be a pure index of the
attentional orienting processes influenced by the presentation of taboo
words.
Nevertheless, the specific results obtained here allow us to use
another baseline, to which the criticism addressed by Mogg et al.
(2008) does not apply. Since attentional biases towards taboo words
were only observed when the taboo word was right-presented, RTs on
left-presented taboo words can serve as a baseline to determine
whether participants were particularly vigilant to right-presented
taboo words or, instead, had difficulties to disengage from these
words. Attentional vigilance to right-presented taboo words would be
reflected in RTs on right-located beeps presented after rightpresented taboo words being shorter than the average RTs on leftand right-located beeps presented after left-presented taboo words.
To put it in a formula: x̅ RTs taboo R-beep R b x̅ RTs taboo L (beep L&R).
In contrast, difficulties to disengage attention from right-presented
taboo words would be reflected in increased RTs on left-located beeps
following right-presented taboo words compared to RTs on beeps
following left-presented taboo words, i.e. x̅ RTs taboo R-beep L N x̅ RTs
taboo L (beep L&R). Given that general, non-spatial effects of taboo
words do not depend on the location of the taboo word in the pair, any
difference between these RTs can only be allocated to spatial effects of
right-presented taboo words.
As can be seen in Fig. 1, in the short interval condition of the present
experiment, RTs on trials in which both the taboo word and the beep
were right-presented were shorter than RTs on trials in which the
taboo word was left-presented (434 vs. 479 ms, t(15) = 2.35,
p = 0.035). In contrast, RTs on trials in which the right-presented
taboo word was followed by a left-located beep did not differ from RTs
on trials in which the taboo word was left-presented (515 vs. 479 ms, t
(15) = 1.75, p = 0.10). Hence, these results suggest that attention
would be engaged preferentially to the spatial location of a rightpresented taboo word relative to the spatial location of the neutral
word presented simultaneously in the pair.
2.3. Discussion
In Experiment 1, participants had to detect a beep as quickly as
possible and to repeat on each trial a word presented at a prespecified side. We observed an attentional bias: the detection of a
beep was faster if it appeared at the same spatial location as a just
presented taboo word than if it appeared at the opposite location. As
predicted, this attentional bias was observed only when the beep
followed immediately the presentation of the pair of words,
suggesting that the phenomenon is very fleeting. Moreover, this
only held true when the taboo word originated from the right side of
the listener, a specificity that allowed us to establish that the observed
attentional bias reflected attentional vigilance towards taboo words
rather than difficulties to disengage from these words.
In addition to producing an attentional bias, taboo words also seem
to lead to a general, non-spatial effect. Indeed, whatever the interval
length this time, taboo pairs yielded longer RTs to the beep than the
other types of pairs, be they emotional or neutral ones.
Contrary to our prediction, there was no attentional bias linked to
the spatial source of negative words. However, in this experiment, the
concurrent repetition task may have monopolized some attentional
resources, making it difficult for negative words to facilitate
selectively the detection of a same-side beep. This explanation was
checked in Experiment 2 by removing the repetition task.
3. Experiment 2: Beep probe detection
This experiment was similar to Experiment 1, except that
participants only had to detect the beep, without having to pay
attention to one side of presentation in order to repeat the presented
word. Moreover, given the results of the first experiment, we only
used the short interval condition for all participants.
In addition to reducing the cognitive load of the task, the present
experiment could also help us to discard an alternative explanation of
the effects observed on taboo words in Experiment 1. Indeed, in the
former experiment, the observed bias and delayed RTs could be linked
to the participants' emotional reaction to the requirement of repeating
a taboo word in front of the experimenter, in a similar way as the taboo
nature of the verbal response made by the subject could be responsible
for the higher recognition thresholds and galvanic skin response
observed for those words (Zajonc, 1962). If this were the case, no
specific effect of taboo words should be observed in the present
experiment. On the contrary, if the effects observed in Experiment 1
reflected attentional influences of taboo words, the same attentional
Please cite this article as: Bertels, J., et al., Emotional valence of spoken words influences the spatial orienting of attention, Acta Psychologica
(2010), doi:10.1016/j.actpsy.2010.02.008
ARTICLE IN PRESS
J. Bertels et al. / Acta Psychologica xxx (2010) xxx–xxx
bias and delayed RTs should be observed here, because the present
situation no longer requires a verbal response.
Table 4
Mean correct detection times observed in Experiment 2 for all types of pairs (standard
errors are in parentheses).
Type of emotional pair
3.1. Method
3.1.1. Participants
Participants were 27 undergraduate students of the Université
Libre de Bruxelles (24 women; 3 left-handed), ranging in age from 18
to 25 (mean: 18.97). They were given course credits for their
participation. The results of three participants (3 women; 1 lefthanded) were discarded from further analyses, because their RTs
were superior to two standard deviations above the mean. All
participants had spoken French for at least 10 years. Average scores
on three personality questionnaires evaluating anxiety and depression levels as well as willingness to be socially desirable are presented
in Table 1.
3.1.2. Material and apparatus
The material and the apparatus were the same as in Experiment 1.
3.1.3. Procedure
The procedure was exactly the same as in Experiment 1 except
that there was no repetition task and we only used the short interval
condition: for all participants on critical pairs, the beep appeared
immediately after the offset of the words.
3.2. Results
Errors (misses and false alarms) were again very infrequent (0.2%,
on the average) so that the analyses focused only on the RT data. As in
Experiment 1, RTs on erroneous trials as well as RTs on 24 pairs were
excluded from the analysis (see Method of Experiment 1).3 Table 4
displayed the mean correct RTs.
As in the previous experiment, we first applied a 4 (type of
emotional pair: neutral/negative/positive/taboo) × 2 (beep location:
left/right) repeated measures analysis of variance (ANOVA) design on
correct response latencies to examine whether the presence of an
emotional word in the pair influenced the processing of an ensuing
beep. The type of emotional pair and the beep location were withinsubject factors. The type of emotional pair was the only significant
factor, F(3, 69) = 3.51, p = 0.02, partial η2 = 0.132. Bonferroni adjusted post-hoc comparisons showed that RTs on taboo pairs (294 ms)
tended to be longer than RTs on negative pairs (286 ms), p = 0.09.
This was not the case for taboo pairs relative to positive ones
(297 ms), p N 0.10.
The same effect was observed in the 3 (type of emotional pair:
negative/positive/taboo) × 2 (beep location: left/right) × 2 (emotional
word location: left/right) repeated measures ANOVA design applied on
correct response latencies, F(2, 46) = 4.41, p b 0.02, partial η22 = 0.161.
Taboo pairs led to significantly longer RTs than negative pairs, p b 0.05.
No difference was observed for taboo pairs compared to positive ones,
p N 0.10 (Bonferroni corrected). No other effect or interaction was
significant. In particular, the three-way interaction between type of
emotional pair, beep location and emotional word location felt short of
significance, F(2, 46) = 2.71, p = 0.078, partial η2 = 0.105. Nevertheless,
Bonferroni adjusted comparisons revealed that, as in Experiment 1,
3
Results from the statistical analyses on the whole set of data were similar to the ones
obtained without the excluded pairs: the three-way interaction between type of emotional
pair, beep location and emotional word location was not significant, F(2, 46) =2.29, pN 0.10,
partial η2 = 0.09. Nevertheless, right-presented beeps following right-presented taboo
words tended to lead to shorter RTs than left-presented beeps following right-presented
taboo words, p= 0.065 (Bonferroni corrected).
7
Beep and emotional word
locations
Neutral
Beep Left
Emotional
Emotional
Beep Right
Emotional
Emotional
Average
297 (12)
word Left
word Right
Negative
Positive
Taboo
292 (11)
282 (11)
301 (15)
294 (14)
294 (13)
303 (14)
285 (13)
286 (12)
286 (12)
291 (14)
301 (13)
297 (14)
295 (14)
286 (12)
294 (13)
294 (11)
word Left
word Right
295 (12)
when the taboo word was right-presented, participants detected the
beep faster when it was also right-presented than when it was leftpresented (286 vs. 303 ms), p = 0.05. No difference was observed when
the taboo word was left-presented, p N 0.10 (see Fig. 2). These
comparisons were not significant for negative and positive pairs, all
p N 0.10. We thus observed an attentional bias linked to taboo words
(17 ms) when these were right-presented. This bias was not correlated
with any of the scores on the personality questionnaires (STAI-state:
r = 0.04; STAI-trait: r = −0.016; BDI–II: r = −0.076; SDS: r = −0.057,
all p N 0.50).
Moreover, the attentional bias linked to right-presented taboo
words was significantly larger than the non-significant bias associated
with right-presented positive words (−7 ms), t(23) = 2.969, p b 0.015
(Bonferroni corrected). However, the difference between the attentional biases linked to right-presented taboo and negative words
(−4 ms) was not significant, p N 0.10.
Given that, as in Experiment 1, an attentional bias was only
observed when the taboo word was right-presented, we used RTs on
trials in which taboo words were left-presented as a baseline in order
to determine which spatial orienting component was affected by the
presentation of a right-presented taboo word. However, RTs on trials
in which both the taboo word and the beep were right-presented did
not significantly differ from RTs on trials in which the taboo word was
left-presented (286 vs. 295 ms, t(23) = 1.36, p N 0.10). Similarly, RTs
on trials in which the right-presented taboo word was followed by a
left-located beep did not differ from RTs on trials in which the taboo
word was left-presented (303 vs. 295 ms, t(23) = 1.16, p N 0.10).
Hence, these results do not allow us to conclude about the nature of
the spatial component affected by the presentation of taboo words.
Fig. 2. Target detection latencies observed for taboo pairs in Experiment 2, as a function
of the emotional word and beep locations.
Please cite this article as: Bertels, J., et al., Emotional valence of spoken words influences the spatial orienting of attention, Acta Psychologica
(2010), doi:10.1016/j.actpsy.2010.02.008
ARTICLE IN PRESS
8
J. Bertels et al. / Acta Psychologica xxx (2010) xxx–xxx
3.3. Discussion
As in Experiment 1, we observed an attentional bias linked to the
taboo nature of words. Taboo pairs also yielded longer RTs than
negative pairs. Given that participants did not now have to repeat
words, the effects observed in Experiment 1 cannot be attributed to an
emotional reaction when repeating a taboo word. The fact that these
effects were larger in Experiment 1 than in the present experiment is
most probably due to the longest RTs associated to the double task
condition of our first experiment. Consistent with this idea, for
Experiments 1 and 2 considered together, we observed a significant
correlation between attentional biases linked to right-presented
taboo words and the average RTs, r = .461, p b 0.01.
Given that no attentional bias was observed for negative words,
contrary to what we supposed, it does not seem that the presence or
absence of such an effect of these words is, at least exclusively, related
to the extent to which attentional resources are mobilized by a
concurrent task. It may be the case that a spatial dimension needs to
be involved, in addition to a low cognitive load, in order to observe
attentional biases towards the spatial location of negative words.
Indeed, as already mentioned in the Introduction of Experiment 1,
some authors argued that spatial attentional orienting towards the
location of a single peripheral cue would only occur when a spatial
representation is activated by the task (Rhodes, 1987; Schmitt et al.,
2000; Spence & Driver, 1994). Supporting this idea, they observed
validity effects (i.e. spatial attentional orienting) with a localization
task but not with a detection task. Considering that the presence of an
emotional word in the pair can act as a single peripheral cue, one
possibility is that attentional orienting towards its spatial location (i.e.
attentional biases) occurs more easily if a localization task is used,
because this task strongly activates a spatial representation. This
possibility was investigated in Experiment 3.
4. Experiment 3: Beep probe localization
This experiment was exactly the same as Experiment 2, except that
participants had to localize the beep when it was presented, instead of
simply detecting it. Word repetition was not required, so that,
contrary to Experiment 1, participants' attentional resources were not
mobilized by a concurrent task. In doing so, we predicted that we
would observe not only attentional biases linked to the taboo nature
of the words (as was the case in the two previous experiments)
but also attentional biases linked to the spatial location of negative
words.
4.1. Method
4.1.1. Participants
Participants were 23 undergraduate students of the Université
Libre de Bruxelles (23 women; 3 left-handed), ranging in age from 18
to 22 (mean: 19.35). They were given course credits for their
participation. The results of three subjects (3 women, 1 left-handed)
were discarded from further analyses because their error rate was
superior to two standard deviations above the mean. All participants
had spoken French for at least 10 years. Average scores on three
personality questionnaires evaluating anxiety and depression levels
as well as willingness to be socially desirable are presented in Table 1.
4.1.2. Material and apparatus
The material and the apparatus were the same as in Experiments 1
and 2.
4.1.3. Procedure
The procedure was exactly the same as in Experiment 2 except
that the participants had to localize the beep as quickly and accurately
Table 5
Mean correct localization times observed in Experiment 3 for all types of pairs (standard
errors are in parentheses).
Type of emotional pair
Beep and emotional
word locations
Neutral
Beep Left
Emotional
Emotional
Beep Right
Emotional
Emotional
Average
358 (18)
word Left
word Right
Negative
Positive
Taboo
349 (18)
359 (19)
359 (19)
360 (21)
374 (23)
384 (22)
352 (16)
335 (16)
349 (17)
329 (15)
354 (18)
350 (18)
367 (20)
348 (19)
368 (20)
349 (17)
word Left
word Right
353 (17)
as possible by pressing the left or right key of a button box. Again,
there was no repetition task.
4.2. Results
As in Experiments 1 and 2, errors (misses, false alarms and wrong
localizations) were very infrequent (0.4%, on the average) so that the
analyses focused only on the RT data. Erroneous responses were
excluded from further analyses. Also, we excluded from the analyses
the RTs on 24 pairs (see Method of Experiment 1).4 Mean correct RTs
are presented in Table 5.
We first applied a 4 (type of emotional pair: neutral/negative/
positive/taboo) × 2 (beep location: left/right) repeated measures
analysis of variance (ANOVA) on response latencies. Both variables
were treated as within-subject factors. The effect of type of emotional
pair was significant, F(3, 57) = 8.57, p b 0.01, partial η2 = 0.311, with
taboo pairs leading to overall longer RTs (368 ms) than neutral
(353 ms), positive (350 ms) and negative (349 ms) pairs, all p b 0.03
(Bonferroni corrected). Also significant was the effect of beep location,
F(1, 19) = 9.48, p b 0.01, partial η2 = 0.333, which reflects the fact that
participants localized the beep faster when it was right-presented
than when it was left-presented (348 vs. 363 ms, respectively).
Further analyses only took the results on emotional pairs into
account. A 3 (type of emotional pair: positive/negative/taboo) × 2
(beep location: left/right) × 2 (emotional word location: left/right)
repeated measures ANOVA design was applied on response latencies.
All variables were within-subjects factors. The main effects that
reached significance were the type of emotional pair, F(2, 38) = 9.32,
p = 0.001, partial η2 = 0.329, and the beep location, F(1, 19) = 9.64,
p b 0.01, partial η2 = 0.336, both reflecting the same differences as in
the analysis including neutral words (see above). The three-way
interaction between the type of emotional pair, beep location and
emotional word location was also significant, F(2, 38) = 6.01,
p = 0.005, partial η2 = 0.24.
To further examine this interaction, we analyzed separately the results
for each type of emotional pair. The interaction between beep location and
emotional word location was significant for the taboo and negative
pairs, F(1, 19)=4.45, pb 0.05, partial η2 =0.19 and F(1, 19)=9.33,
pb 0.01, partial η2 =0.329, not for the positive pairs, F(1, 19)=2.92,
pN 0.10, partial η2 =0.133. The interaction between beep location and
emotional word location for taboo pairs reflected the fact that when
the taboo word was right-presented, participants localized a right-
4
Note that results from the statistical analyses including the excluded pairs were similar
to those obtained without these: the three-way interaction between type of emotional
pair, beep location and emotional word location is significant, F(2, 38) = 5.25, p = 0.01,
partial η2 = 0.216.
Please cite this article as: Bertels, J., et al., Emotional valence of spoken words influences the spatial orienting of attention, Acta Psychologica
(2010), doi:10.1016/j.actpsy.2010.02.008
ARTICLE IN PRESS
J. Bertels et al. / Acta Psychologica xxx (2010) xxx–xxx
presented beep faster (348 ms) than a left-presented beep (384 ms),
p=0.004. There was no difference when the taboo word was leftpresented, pN 0.10 (see Fig. 3a). This was also the case for negative
pairs: when the negative word was right-presented, participants
localized a right-presented beep faster (335 ms) than a left-presented
one (359 ms), p=0.002, while there was no difference when the word
was left-presented, pN 0.10 (see Fig. 3b).
We thus observed attentional biases linked to taboo (36 ms) and to
negative (24 ms) words when they were right-presented. None of these
attentional biases was correlated with any of the scores on the
personality questionnaires (For taboo words: STAI-state: r = 0.07,
STAI-trait: r = 0.23, BDI-II: r = −0.13, SDS: r = −0.10, all p N 0.30; for
negative words: STAI-state: r = −0.06, STAI-trait: r = 0.15, BDI–II:
r = 0.21, SDS: r = −0.30, all p N 0.15 ).
The difference between the attentional biases linked to rightpresented negative and taboo words and the attentional bias linked to
right-presented positive words (6 ms) did not reach significance, both
p ≥ 0.10 (Bonferroni corrected). As expected, the biases linked to rightpresented negative and taboo words did not differ from each other, t b 1.
As can be seen in Fig. 3a, RTs on trials in which both the taboo word
and the beep were right-presented were shorter than RTs on trials in
which the taboo word was left-presented (348 vs. 370 ms, t(19) =
2.34, p = 0.03). In contrast, RTs on trials in which the right-presented
taboo word was followed by a left-located beep did not differ from RTs
on trials in which the taboo word was left-presented (384 vs. 370 ms,
t(19) = 1.2, p N 0.10). Similarly, as depicted by Fig. 3b, RTs on trials in
which both the negative word and the beep were right-presented
were shorter than RTs on trials in which the negative word was leftpresented (335 vs. 350 ms, t(19) = 2.34, p = 0.03). On the contrary, no
RT difference was found between trials in which a left-located beep
Fig. 3. a and b. Target localization latencies observed in Experiment 3, as a function of
the emotional word and beep locations, for taboo (a) and negative (b) pairs,
respectively.
9
followed a right-presented negative word and trials in which a negative
word was left-presented (359 vs. 335 ms, t(19) = 1.194, p N 0.10).
4.3. Discussion
As in Experiments 1 and 2, we observed that participants
responded faster to a beep presented at the same location as a
taboo word than presented at the same location as a neutral word.
Again, this attentional bias was only observed when the taboo word
was right-presented. Moreover, as in Experiment 1, results suggest
that participants did not have difficulties in disengaging attention
from taboo words but were, rather, particularly vigilant to these
words. This spatial effect of taboo words thus appears to be robust. In
addition, as in the previous experiments, trials in which a taboo word
was presented led to overall longer RTs to a following beep than other
types of trials, whatever the side of presentation of the beep.
Importantly, in Experiment 3, an attentional bias was also
observed towards negative words. As with the attentional biases
observed for taboo words, this bias for negative words was only
observed when the emotional word was right-presented and most
probably reflects attentional vigilance towards these words.
5. General discussion
While the distinction between the “what” and the “where” auditory
subsystems is largely accepted (e.g., Kubovy & Van Valkenburg,
2001), the functional link between these subsystems has been
relatively neglected. Yet such a link would be advantageous for
ecological adaptation: if the identification of an auditory object that
conveys emotional information leads to activate the information about
its spatial location, then one would be in a better attentional state to
react to subsequent auditory information coming from the same spatial
location.
In the present study, we investigated the existence of a functional
link between the auditory “what” and “where” subsystems by
examining the influence of the emotional content of spoken words,
linked to the “what” subsystem, on the spatial orienting of auditory
attention, relying on the “where” subsystem. This was done by using a
specific kind of cuing situation (Posner, 1980), actually an auditory
adaptation of the attentional deployment task (MacLeod et al., 1986)
in which a pair of spoken words — one potentially emotional — was
followed by a beep target. Considering that the presence of an
emotional word of the pair can act as a single peripheral cue, we
expected to observe spatial congruency effects, namely attentional
biases linked to the spatial location of taboo and negative words, but
no specific effect of positive words was anticipated. Moreover,
according to Kinsbourne's (1970) theory, as the presentation of
verbal material could prime the left hemisphere, leading to an
orientational bias to the contralateral hemifield, and following the
idea that an emotional word catches more attentional resources than
its paired neutral word (therefore activating to a greater extent the
hemispheric contralateral to its side of presentation), we predicted
attentional biases linked to the emotional valence of the words to
occur largely when the emotional word was right-presented.
An overview of the experiments and their associated results is
presented in Table 6. Crucially, in addition to non-spatial inhibitory
effects of taboo words, attentional biases linked to the spatial location
of taboo and negative spoken words were observed. Namely,
participants responded faster to a beep presented at the same spatial
location as the taboo or negative word of the pair than to the neutral
word. In particular, attentional biases towards taboo words were
consistently observed in three experiments, regardless of whether
attention to the words and a spatial response were required or not. In
contrast, attentional biases towards negative words were only
observed when the attentional load was low, that is when attention
to the words was not required, and when participants had to indicate
Please cite this article as: Bertels, J., et al., Emotional valence of spoken words influences the spatial orienting of attention, Acta Psychologica
(2010), doi:10.1016/j.actpsy.2010.02.008
ARTICLE IN PRESS
10
J. Bertels et al. / Acta Psychologica xxx (2010) xxx–xxx
Table 6
Overview of the three experiments.
Experiment Interval
Task(s)
length (ms)
Results
Spatial effects
Non-spatial effects
(attentional biases)
1
0
250
2
0
3
0
Beep
detection
then
repetition
Beep
detection
then
repetition
Beep
detection
Beep
localization
Right taboo words
RTs taboo N RTs
negative, positive,
neutral words
–
RTs taboo N RTs
negative, positive,
neutral words
Right taboo words
RTs taboo N RTs
negative words
RTs taboo N RTs
negative, positive,
neutral words
Right taboo and
negative words
emotional trials exclusively. Hence, if there were any influence of
the predictive nature of the occurrence of an emotional word, this
influence marred all these reaction times, making them comparable
and warranting that any difference between them is due to their
specific emotional content.
Finally, in Experiment 3, although target beeps were also
presented more often after emotional than after neutral trials, the
participants' task was to localize the beep. Thus, an adequate response
could not be prepared on the basis of the occurrence of an emotional
word in the pair, given that the location of the emotional word of the
pair (left/right) was not predictive of the location of the beep. Still, we
observed results similar to those of Experiments 1 and 2, indicating
that the predictability of the relation between the presence of an
emotional word and the occurrence of the beep, even if it were noted
by the participants (which is unwarranted), could not be responsible
for the observation of spatial and non-spatial effects for taboo and
negative words, respectively.
5.1. Attentional biases to taboo words
the spatial location of the beep by pressing one of two response
buttons (as in Experiment 3). Importantly, these attentional biases
occurred only when the emotional word of the pair was rightpresented, which is coherent with Kinsbourne's (1970) theory.
Notably, these attentional biases were observed in unselected
volunteers and were not correlated with any of the scores in the
personality questionnaires. Our data are thus in accordance with the
idea of an emotion-dependent functional link between the “what” and
the “where” auditory subsystems.
Before discussing the conditions under which this link is observed, a
relevant methodological point should be addressed. In all experiments,
we adopted MacLeod et al.'s (1986) original design of the dot probe task
that was largely used in other visual studies (e.g., Fox, 1993; Salemink
et al., 2007). As previously commented on by Mogg and Bradley (1998,
1999), in this design, the occurrence of an emotional word is predictive
of the occurrence of the target. More specifically, in the present
experiments the beep was presented after all emotional pairs but only
after one third of the neutral pairs.
Nevertheless, the present results cannot be accounted for by the
predictability of the occurrence of the beep. As a matter of fact, in the
present study, we used several types of emotional words (i.e., negative,
positive, and taboo). Thus, if the predictability of the occurrence of the
beep following an emotional word were responsible for the observed
spatial and non-spatial effects, then these effects would be observed for
all emotional words. On the contrary, when a detection task was used
(Experiments 1 and 2), only taboo words influenced the spatial
orienting of attention, only negative words led to non-spatial effects,
and positive words did not elicit either type of effect. Moreover,
attentional biases were only observed when the emotional word was
right-presented in the linguistic pair. This cannot be a consequence
of the fact that the presence of an emotional word in the pair was
predictive of the occurrence of the beep, given that the side of presentation of the emotional word was unpredictable. Indeed, it does not
seem reasonable to admit that listeners only attended to taboo words on
their right side or that they decided that only taboo words presented on
their right side were predictive of the occurrence of the beep. However,
even if the present results eliminate the possibility that the attentional
effects would result from the detection of the predictability of the
occurrence of the target, further studies should rather present a to-belocalized beep on each (emotional or neutral) trial, as it is often the case
in dot probe studies (e.g., Mogg, Bradley, De Bono, & Painter, 1997; Mogg
& Bradley, 1999).
It ought also to be taken into account that the occurrence of
attentional biases linked to the presentation of a specific type of
emotional word was investigated comparing reaction times to
An attentional bias linked to the spatial source of right-presented
taboo words was consistently observed in three experiments. This is
consistent with previous results from visual studies that used the
taboo variant of the emotional Stroop task (the taboo Stroop task), in
which it has been repeatedly shown that healthy, unselected
participants took longer to name the ink colour of taboo words
compared to neutral words (MacKay & Ahmetzanov, 2005; MacKay
et al., 2004; Siegrist, 1995; Taylor, Kornblum, Lauber, Minoshima, &
Koeppe, 1997). These interference effects are generally considered to
be attentional biases. However, as in the Stroop setting the two
dimensions of the stimulus, relevant and irrelevant for the task (the
ink colour and the emotional valence, respectively), are part of the
same stimulus, no spatial attentional shift is expected to occur. Given
that taboo words have never been used in paradigms involving spatial
attentional orienting such as the dot probe task (MacLeod et al., 1986)
and the emotional cuing paradigm (Stormark, Nordby, & Hugdahl,
1995), the present results extend the findings already reported in the
Stroop situation for visually presented taboo words by showing
spatial attentional biases for auditorily presented taboo words. In
addition, in comparison to most visual studies using taboo words (but
see Nielsen & Sarason, 1981), the present results may have more
ecological validity, given that such words are more common in oral
than in written language.
We also examined, more specifically, on which spatial attentional orienting component(s) taboo words had an effect. For this
purpose, we compared RTs on congruent and incongruent rightpresented taboo trials with RTs on left-presented taboo trials
(acting as baseline, given that no attentional bias was observed on
these trials). In Experiments 1 and 3, we observed an RT advantage
for congruent right-presented taboo trials compared to baseline
left-presented taboo trials, but found no difference between RTs on
incongruent right-presented taboo trials and left-presented taboo
trials. This is more consistent with the hypothesis of an attentional
vigilance towards taboo words than with the idea of difficulties to
disengage attention from these words once attention has been
normally engaged. Attention would be engaged preferentially to
the spatial location of a right-presented taboo word relative to the
spatial location of the neutral word presented simultaneously in
the pair. However, further studies using a different methodology
should confirm these findings. Indeed, as suggested by an
anonymous Reviewer, the fact that no attentional bias was
observed for left-presented taboo words does not ensure that
they did not have any influence on RTs (e.g., they could be
antagonistic). As such, RTs on emotional trials do not appear to be
the most proper baseline.
Please cite this article as: Bertels, J., et al., Emotional valence of spoken words influences the spatial orienting of attention, Acta Psychologica
(2010), doi:10.1016/j.actpsy.2010.02.008
ARTICLE IN PRESS
J. Bertels et al. / Acta Psychologica xxx (2010) xxx–xxx
Further work should aim at clarifying the basis of this attentional
vigilance towards spoken taboo words. Nevertheless, it is worth
noting that the set of taboo words we used in the experiments was
matched in literary and web frequencies with the set of neutral words
with which they were associated in the pairs. The infrequent nature of
taboo words is therefore unlikely to be responsible for the occurrence
of the attentional bias. Moreover, one of the post-hoc control studies
confirmed that our taboo words were not less familiar than the
associated neutral words (see Method section of Experiment 1).
Rather, two potential factors could have held the participants'
attention: the arousing and shocking nature of the taboo words.
According to various studies, arousal would account for the
observed effects on attention (e.g., Anderson, 2005; Schimmack,
2005) as well as on memory (e.g., Buchanan, Etzel, Adolphs, & Tranel,
2006; Kensinger & Corkin, 2003; Mather et al., 2006). Using the
attentional blink paradigm, recent studies showed preferential
attentional processing of taboo stimuli (Arnell, Killman, & Fijavz,
2007; Mathewson, Arnell, & Mansfield, 2008; Most, Smith, Cooter,
Levy, & Zald, 2007), and some researchers argued that the arousal
nature of these stimuli in fact influences their encoding in memory
(Arnell et al., 2007; Mathewson et al., 2008). However, the whole set
of data does not seem to support the idea that arousal is the only
factor leading to attentional capture: in Mathewson et al.'s (2008)
study, the threatening words used in the experiments were as
arousing as taboo words, but had no effect. Hence, even if arousal is a
crucial dimension, taboo words seem to have a particular quality
which accounts for this different effect.
Although our study was not initially designed to assess the impact
of arousal, two results of our post-hoc control study (see Method
section of Experiment 1) enable us to discard this factor as the main
basis of the observed effects. Firstly, no correlation was found
between the attentional bias when the emotional word was rightpresented and the difference in arousal between the emotional and
the neutral word of the pair in any of the experiments (all p N 0.10).
Hence, this difference does not seem to be a good predictor of the size
of the attentional bias (highest R2: 0.041). This was the case both
when the correlation analyses included all emotional words irrespective of valence and when separate correlation analyses were
performed for positive, negative and taboo words. Secondly, the
arousal level of taboo words was lower than the arousal level of the
negative words used in the experiments. Also, the difference in
arousal between the emotional and the neutral word of the pair was
lower for taboo than for negative pairs. If arousal were the crucial
factor in the occurrence of the effect, we should have observed robust
attentional biases towards negative words as well. Thus, rather than
arousal, it seems that it is their shocking quality (perhaps combined
with arousal) that is the relevant factor of the attentional bias they
provoke. As MacKay et al. (2004) suggested, the particular type of
emotional reaction could play a crucial role in the culmination of the
process. For example, taboo words may create sexual arousal or
embarrassment, which would be responsible for this particular
involvement of attentional processes.
Attentional biases linked to taboo words appear to be shortlived, suggesting that the activation of the “what-where” connections would be transient. As a matter of fact, in Experiment 1,
attentional biases to taboo words were only observed when the
interval between the offset of the words and the onset of the beep
was null, not when it was 250 ms long. The fact that an attentional
bias only occurs towards events (here, the beep) which are
concomitant (e.g., in the emotional Stroop task) or which
immediately follow the emotional information (taboo, in our
case) seems adaptive: given their transitory nature, emotional
auditory stimuli would create reactions having an influence on the
processing of immediately subsequent events, not on the processing of later ones. This would be the case particularly when
subsequent events are not emotional and when they have no other
11
link with the emotional information than spatial location. However,
as the effect of two intervals were compared only in Experiment 1
(given that no attentional bias was observed with the long one)
and as no other study using an attentional deployment task
compared different conditions of intervals between the pair of
words and the probe, our observation remains isolated, and hence
should be considered with caution.
5.2. Attentional biases towards negative words
In addition to attentional biases to right-presented taboo words,
we also observed attentional biases to right-presented negative words
in Experiment 3, but this was not the case in Experiments 1 and 2.
Therefore, the attentional bias to negative words appears to be less
robust than the attentional biases to taboo words or, instead, to occur
under specific conditions.
A crucial factor could be cognitive load, with spatial effects of the
negative valence observable only under low-load conditions. Indeed,
in Experiment 3, the average RTs were far shorter than in
Experiment 1 (344 vs. 439 ms). The idea that task-irrelevant
emotional stimuli are subject to the effects of processing load (as
are neutral stimuli, cf. Lavie, 1995) is supported by studies showing
that brain responses usually observed to emotional stimuli disappear
when cognitive load increases (e.g., Doallo, Holguín, & Cadaveira,
2006; Mitchell et al., 2007; Pessoa, McKenna, Gutierrez, & Ungerleider, 2002; Pessoa, Padmala, & Morland, 2005). Also, at the
behavioural level, interference effects of emotional stimuli on
performance vanish with the increase of task demands, suggesting
that interference only occurs when sufficient resources are available
to process task-irrelevant emotional information (Okon-Singer,
Tzelgov, & Henik, 2007; Smith Erthal et al., 2005). Similarly, in our
study, participants were particularly vigilant to negative stimuli
when cognitive load was low (Experiment 3), not at all when it was
high (Experiment 1).
However, cognitive load does not explain the whole phenomenon.
As a matter of fact, no attentional bias to negative words was observed
in Experiment 2, although participants only had to detect the presence
of a beep. As already developed in the Discussion section of
Experiment 2, one possibility is that negative words would only
influence the spatial orienting of auditory attention when the spatial
location of the beep is task-relevant, as was the case in Experiment 3,
but not in Experiment 2. Actually, when negative words are
presented, the involvement of a spatial orienting component would
be critical to permit “what”-related dimensions to influence “where”related ones, in addition to a low cognitive load. The “where”
subsystem would be influenced by the “what” subsystem that detects
a negative emotional content only when it is previously activated,
namely when it is somehow relevant to the participant's goal. These
assumptions require verification in further studies.
Interestingly, as was the case with taboo words, attentional biases
to right-presented negative words would be due to participants being
particularly vigilant to these words: when presented together with a
neutral word, attention would be oriented preferentially to the spatial
location of the negative word.
5.3. Laterality effects in attention
Following Kinsbourne's (1970) theory, we assumed that the use of
verbal material would prime the left hemisphere, leading to
orientational biases to the contralateral side (the right one).
Ordinarily, one would expect left rather than right orientational
biases, given the emotional nature of the words and the fact that
emotions are processed in the right hemisphere (e.g., Bryden &
MacRae, 1989). However, this assumption comes from studies in
which participants had to process the emotional dimension of the
stimuli, which was not the case in our experiments. In Experiments 2
Please cite this article as: Bertels, J., et al., Emotional valence of spoken words influences the spatial orienting of attention, Acta Psychologica
(2010), doi:10.1016/j.actpsy.2010.02.008
ARTICLE IN PRESS
12
J. Bertels et al. / Acta Psychologica xxx (2010) xxx–xxx
and 3 they were not even required to process the words at all.
Moreover, not all words were emotionally charged. Hence, the mere
presentation of words seems to activate the left hemisphere, and this
activation could have been reinforced by the expectations of the
participants as to the verbal nature of the major part of the material,
leading to a right orientational bias. This idea resonates with the
earlier findings of Demakis and Harrison (1994) who showed that the
rehearsal of affective words primed the left hemisphere, as neutral
words did, suggesting that even affective words are processed by the
left hemisphere. Moreover, Russell and Voyer (2004) showed that the
magnitude of the REA is similar in studies where an emotional
component was present and in studies where it was not. Finally, it is
worth noting that, according to Bryden (1978) and Cheatham and
Herbig (1992), there is a generalized rightward orientational bias in
right-handed people.
As indicated in the Introduction, we predicted the processing of
right-presented congruent beeps to benefit not only from the
attraction of attentional resources to the spatial source of the
emotional word (which would be the case also of left-presented
congruent beeps), but also from the left-hemisphere activation
associated with the recurrent processing of verbal material. The
results of these three experiments confirmed this prediction.
Moreover, only right-presented emotional words influenced significantly the spatial orienting of attention, and consequently the
processing of the beep. Thus, the emotional valence of the words
only has a systematic (in the case of taboo words) and reliable
influence if it is presented at the side of the auditory space primed
by left-hemisphere activation, hence locking the connections
between the “what” and “where” systems only in the right side of
the space.
This seems to constitute a partial failure of evolutionary
adaptation. It could indeed be so. Non-adaptive behaviours could
in some circumstances result from the conjunction of two or more
mechanisms that are, by themselves, adaptive. The adaptive value
of hemisphere specialisation, particularly for language, is largely
recognized (for recent reviews, see e.g., Hopkins, 2008; Josse &
Tzourio-Mazoyer, 2004). This is also the case for the anatomical
and functional decussation of the central nervous system relating,
mostly in the vertebrates, each half of the brain to the opposite
half of the body and of the environmental space, although its
precise adaptive origin has been matter for debate for many years
(see, e.g., Ramon y Cajal, 1898, and Suttie, 1926, vs. Braitenberg,
1965, for quite opposite views). The combination of two distinct
adaptive mechanisms may not always be adaptive itself, such as
when an auditory non-verbal stimulus is presented on the
opposite side of a previous word (it being particularly relevant
for some reason) which is processed in the left hemisphere and
engages attention to the right of environmental space. Moreover,
the asymmetrical pattern of attentional biases observed here could
also depend on the specific experimental setting that we designed.
If one considers sounds to be usually delivered by an object, and
that they serve to attract visual attention towards this object, our
experiments may have lacked ecological validity to some extent. A
cross-modal setting, with an auditory emotional word followed by
a visual target, could not lock the “what”–“where” connections
exclusively to the right side and might then give rise to a fully
adaptive attentional behaviour. Such experiments are currently in
progress.
In comparison to the unimodal setting used here, the cross-modal
setting will offer a further advantage in allowing reducing the SOA
(more precisely, the word-beep interval) without masking (through
noise effects) the spoken words. This will allow better comparison
with studies investigating attentional biases in the visual modality,
since written words are generally briefly presented. In addition,
observations would not be limited to a relatively late and limited
snapshot of the influence of emotional stimuli on attentional orienting.
5.4. Non-spatial effects of emotional words
In all experiments, taboo pairs led to longer RTs to the beep. These
inhibitory effects of taboo words ought to be considered as nonspatial, because (a) they do not depend on the spatial congruence
between the presentation sides of the emotional word and the beep
and (b) they are observed even when no spatial dimension is involved
(Experiment 2), namely when participants did not have to attend one
side of space in order to repeat a word (as in Experiment 1) nor to
localize the beep on each trial (as in Experiment 3).
These effects can be related to the interference effects of taboo
words observed in taboo Stroop tasks (e.g., MacKay & Ahmetzanov,
2005) and to their slowing effects in lexical decision tasks (Thomas &
LaBar, 2005; but see MacKay et al., 2004), and interpreted as a freezing
reaction, namely an inhibition of motor responses provoking a generic
slowdown, similar to the interpretation of the interference and
slowing effects of negative words (Algom, Chajut, & Lev, 2004). The
fact that these non-spatial effects were also observed when there was
a 250 ms delay between the offset of the words and the onset of the
beep (long interval condition of the Experiment 1) supports this
interpretation: indeed, Flykt (2006) suggested that if an effect of a
threatening stimulus is still present after detection of this stimulus
(namely during the withholding of the response), this effect can be
located at the level of the response. Alternatively, some authors
consider that general non-spatial effects of emotional words would
actually reflect difficulties to disengage attention from their taskirrelevant emotional content (e.g., Estes & Verges, 2008; McKenna &
Sharma, 2004) or the triggering of binding mechanisms which
associate the source of an emotion (here, the word) with its context
of occurrence (see MacKay et al., 2004 and MacKay et al., 2005).
Further studies should aim at disentangling these interpretations and
identifying the basis of the non-spatial effects we observed.
For the time being, the fact that for taboo words general nonspatial effects were observed in both the short and the long interval
conditions of Experiment 1 but attentional biases only in the short
interval condition, indicates that these two effects are distinct and
independent from each other. The general non-spatial effects are
more long-lasting and less dependent on the immediate temporal
succession of the emotional stimulus and the beep than the spatial
attentional biases are, at least for taboo words.
5.5. Conclusion
The present data are in agreement with the idea that the functional
link between the “what” and the “where” auditory subsystems is
reinforced by the emotional characteristics of the stimuli. The
identification of a spoken emotional word automatically activates
information about its spatial location, influencing the processing of a
subsequent extraneous auditory stimulus. Nevertheless, only emotional words presented at the automatically primed side of the
auditory space (in our case, the right side) affected the orienting of
attention. This held true for taboo and, to a lesser extent, for negative
words. This suggests that the “what”–“where” connections are locked
exclusively to the right side of the space in a purely auditory situation
using a linguistic material. Further work will aim to assess whether
such a limitation disappears in audio-visual settings or with nonlinguistic emotional auditory materials.
Acknowledgements
This work was supported by two grants of the F.R.S.–FNRS
(convention FRFC 2.4579.02 F and FNRS 1.5705.06). Many thanks to
Marcha Van Boven for the help in preparing the material, to Olivia
Gosseries and Ana Franco for testing the participants, as well as to
Pascale Lidji and Chotiga Pattamadilok for their technical assistance.
Please cite this article as: Bertels, J., et al., Emotional valence of spoken words influences the spatial orienting of attention, Acta Psychologica
(2010), doi:10.1016/j.actpsy.2010.02.008
ARTICLE IN PRESS
J. Bertels et al. / Acta Psychologica xxx (2010) xxx–xxx
Appendix A
Appendix A (continued)
Emotional word
French word
Words constituting the emotional pairs (included and excluded from the statistical
analyses on the basis of the post hoc control studies), and their English translation.
Emotional word
French word
13
Neutral word
English translation
Negative pairs
Included in the statistical analyses
Attaque
Attack
Bombe
Bomb
Bourreau
Torturer
Cadavre
Corpse
Cancer
Cancer
Carnage
Slaughter
Cercueil
Coffin
Chagrin
Grief
Colère
Anger
Coupable
Culprit
Crime
Crime
Danger
Danger
Echec
Failure
Haine
Hate
Hurler
Scream
Malheur
Misfortune
Massacre
Massacre
Menace
Threat
Menottes
Handcuffs
Meurtre
Murder
Misère
Misery
Morgue
Morgue
Mort
Death
Otage
Hostage
Panique
Panic
Peur
Fear
Prison
Jail
Ravage
Devastation
Sanglot
Sob
Seringue
Syringe
Sévices
Abuse
Sida
Aids
Souffrir
Suffer
Tank
Tank
Tombe
Grave
Torture
Torture
Tragique
Tragic
Violence
Violence
Excluded from the analyses
Famine
Starvation
Hargne
Aggressiveness
Positive pairs
Included in the statistical analyses
Amour
Love
Beauté
Beauty
Bon
Good
Cadeau
Gift
Câlin
Cuddle
Chance
Luck
Charme
Charm
Confiance
Trust
Confort
Comfort
Content
Glad
Couple
Couple
Courage
Courage
Extase
Ecstasy
Fête
Party
Gentil
Kind
Guérir
Recover
Intime
Intimate
Joie
Joy
Merci
Thanks
Parfait
Perfect
Passion
Passion
Plage
Beach
Respect
Respect
French word
English translation
Allure
Bulle
Baigner
Convaincre
Cadence
Corniche
Citrouille
Charbon
Quarante
Quatorze
Cloche
Dessin
Etoffe
Hall
Hardi
Mélange
Moutarde
Minute
Mésange
Muscle
Message
Merle
Mise
Augure
Pendule
Poche
Plateau
Régate
Sourcil
Cerise
Centime
Semer
Secteur
Taupe
Tâche
Tartine
Transit
Voiture
Speed
Bubble
Bathe
Convince
Rhythm
Cornice
Pumpkin
Coal
Forty
Fourteen
Bell
Drawing
Cloth
Hall
Bold
Blend
Mustard
Minute
Chickadee
Muscle
Message
Blackbird
Putting
Oracle
Pendulum
Pocket
Tray
Regatta
Eyebrow
Cherry
Cent
Sow
District
Mole
Task
Slice of bread
Transit
Car
Fissure
Harpe
Crack
Harp
Agence
Bateau
Bas
Coton
Caddie
Chaise
Chiffre
Conduire
Canard
Carré
Course
Commune
Estrade
Phase
Genou
Garage
Indice
Juin
Montrer
Poster
Panier
Presse
Rentrer
Agency
Boat
Low
Cotton
Trolley
Chair
Digit
Drive
Duck
Square
Race
Town
Platform
Phase
Knee
Garage
Clue
June
Show
Post
Basket
Press
Return
(continued on next page)
Neutral word
English translation
Positive pairs
Included in the statistical analyses
Rêve
Dream
Rire
Laugh
Romance
Romance
Sincère
Sincere
Sourire
Smile
Succès
Success
Tendre
Tender
Trésor
Treasure
Vacances
Holiday
Vie
Life
Excluded from the analyses
Argent
Money
Désir
Desire
Gagnant
Winner
Gaieté
Cheerfulness
Partage
Share
Piscine
Swimming pool
Santé
Health
Taboo pairs
Included in the statistical analyses
Bander
Have a hard-on
Bite
Dick
Blondasse
Dull blond
Bonasse
Hottie
Bonniche
Skivvy
Bordel
Brothel
Catin
Trollop
Chiant
Damn
Clodo
Bum
Connasse
Bitch
Conne
Twat
Crapule
Scoundrel
Crétin
Moron
Débile
Moronic
Encule
Bugger
Foireux
Yellow-bellied
Foutre
Cum
Gerber
Throw up
Gueule
Trap
Ignare
Ignorant
Merde
Shit
Merdeux
Squirt
Morveux
Jerk
Nichon
Tit
Niquer
Fuck
Pédé
Queer
Pétasse
Slut
Péter
Fart
Pine
Cock
Pisser
Pee
Pouffiasse
Tart
Putain
Whore
Roulure
Trollop
Salope
Bitch
Tapette
Queer
Taré
Cretin
Taulard
Con
Traînée
Slut
Troncher
Shag
Excluded from the analyses
Biroute
Prick
French word
English translation
Robe
Rôle
Raquette
Salade
Seconde
Sûrement
Tourne
Trafic
Visite
Va
Dress
Role
Racket
Salad
Second
Surely
Turn
Traffic
Visit
Go
Aspect
Décor
Gazon
Galop
Parcours
Pilote
Salon
Appearance
Decor
Lawn
Gallop
Way
Pilot
Lounge
Bambou
Bêche
Brasseur
Balise
Banquise
Bourgogne
Coupon
Chiot
Cresson
Canine
Quinte
Crevette
Croquis
Décade
Amphore
Foison
Firme
Jongler
Gare
Hippique
Marque
Matelot
Merlan
Napper
Neiger
Piquet
Pudding
Pommeau
Pull
Pigeon
Pochoir
Panneau
Rameur
Serrure
Tisseur
Tympan
Teckel
Trousseau
Tréma
Bamboo
Spade
Brewer
Beacon
Ice field
Burgundy
Coupon
Puppy
Cress
Canine
Quintain
Shrimp
Sketch
Decade
Amphora
Abundance
Firm
Juggle
Station
Horse
Brand
Sailor
Whiting
Coat
Snow
Peg
Pudding
Pommel
Pullover
Pigeon
Stencil
Panel
Rower
Lock
Weaver
Eardrum
Dachshund
Bunch
Dieresis
Benzène
Benzene
References
Algom, D., Chajut, E., & Lev, S. (2004). A rational look at the emotional Stroop
phenomenon: A generic slowdown, not a Stroop effect. Journal of Experimental
Psychology: General, 133(3), 323−338.
Anderson, A. K. (2005). Affective influences on the attentional dynamics supporting
awareness. Journal of Experimental Psychology: General, 134(2), 258−281.
Arnell, K. M., Killman, K. V., & Fijavz, D. (2007). Blinded by emotion: Target misses
follow attention capture by arousing distractors in RSVP. Emotion, 7(3), 465−477.
Please cite this article as: Bertels, J., et al., Emotional valence of spoken words influences the spatial orienting of attention, Acta Psychologica
(2010), doi:10.1016/j.actpsy.2010.02.008
ARTICLE IN PRESS
14
J. Bertels et al. / Acta Psychologica xxx (2010) xxx–xxx
Bar-Haim, Y., Lamy, D., Pergamin, L., Bakermans-Kranenburg, M. J., & van Ijzendoorn, M. H.
(2007). Threat-related attentional bias in anxious and non-anxious individuals: A
meta-analytic study. Psychological Bulletin, 133(1), 1−24.
Baumeister, R. F., Bratslavsky, E., Finkenauer, C., & Vohs, K. D. (2001). Bad is stronger
than good. Review of General Psychology, 5, 323−370.
Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Manual for the Beck Depression Inventory
(2ème édition). San Antonio, TX: The Psychological Corporation.
Bertels, J., Kolinsky, R., & Morais, J. (2009). Norms of emotional valence, arousal, threat
value and shock value for 80 spoken French words: Comparison between neutral
and emotional tones of voice. Psychologica Belgica, 49(1), 19−40.
Bonin, P., Méot, A., Aubert, L., Malardier, N., Niedenthal, P., & Capelle-Toczek, M. -C.
(2003). Normes de concrétude, de valeur d'imagerie, de fréquence subjective et de
valence émotionnelle pour 866 mots. L'Année Psychologique, 104, 655−694.
Braitenberg, V. (1965). Taxis, kinesis and decussation. Progress in Brain Research, 17,
210−222.
Bryden, M. P. (1978). Strategy effects in the assessment of hemispheric asymmetry. In
G. Underwood (Ed.), Strategies of information processing. London: Academic Press.
Bryden, M. P., & MacRae, L. (1989). Dichotic laterality effects obtained with emotional
words. Neuropsychiatry, Neuropsychology, and Behavioral Neurology, 1(3), 171−176.
Buchanan, T. W., Etzel, J. A., Adolphs, R., & Tranel, D. (2006). The influence of autonomic
arousal and semantic relatedness on memory for emotional words. International
Journal of Psychophysiology, 61(1), 26−33.
Buchtel, H. A., & Butter, Ch. M. (1988). Spatial attentional shifts. Implications for the role
of polysensory mechanisms. Neuropsychologia, 26(4), 499−509.
Buchtel, H. A., Butter, Ch. M., & Ayvasik, B. (1996). Effects of stimulus source and intensity
on covert orientation to auditory stimuli. Neuropsychologia, 34(10), 979−985.
Cheatham, B. H., & Herbig, R. (1992). Spatial localization and auditory lateralization:
Binaural cues and their absence. Perceptual and Motor Skills, 75(3), 919−922.
Clarke, S., Bellmann, A., Meuli, R. A., Assal, G., & Steck, A. J. (2000). Auditory agnosia and
spatial deficits following left hemispheric lesions: Evidence for distinct processing
pathways. Neuropsychologia, 38, 797−807.
Cohen, J., MacWhinney, B., Flatt, M., & Provost, J. (1993). Psyscope — An interactive
graphic system for designing and controlling experiments in the Psychology
laboratory using Macintosh computers. Behavior Research Methods, Instruments &
Computers, 25(2), 257−271.
Crowne, D. P., & Marlowe, D. (1960). A new scale of social desirability independent of
psychopathology. Journal of Consulting Psychology, 24(4), 349−354.
Demakis, G. J., & Harrison, D. W. (1994). Subvocal rehearsal of neutral and affective
words interferes with left-hemisphere performance and facilitates right-hemisphere performance. Psychobiology, 22(3), 238−243.
Deutsch, D., & Roll, P. (1976). Separate “what” and “where” decision mechanisms in
processing a dichotic tonal sequence. Journal of Experimental Psychology: Human
Perception and Performance, 2(1), 23−29.
Doallo, S., Holguín, S. R., & Cadaveira, F. (2006). Attentional load affects automatic
emotional processing: Evidence from event-related potentials. NeuroReport, 17
(17), 1797−1801.
Estes, Z., & Verges, M. (2008). Freeze or flee? Negative stimuli elicit selective
responding. Cognition, 108, 557−565.
Flykt, A. (2006). Preparedness for action: Responding to the snake in the grass.
American Journal of Psychology, 119(1), 29−43.
Fox, E. (1993). Allocation of visual attention and anxiety. Cognition and Emotion, 7(2),
207−215.
Goldring, J. E., Dorris, M. C., Corneil, B. D., Ballantyne, P. A., & Munoz, D. P. (1996).
Combined eye-head gaze shifts to visual and auditory targets in humans.
Experimental Brain Research, 111(1), 68−78.
Grühn, D., Smith, J., & Baltes, P. B. (2005). No aging bias favouring memory for positive
material: Evidence from a heterogeneity-homogeneity list paradigm using
emotionally toned words. Psychology and Aging, 20(4), 579−588.
Hafter, E. R. (1997). Binaural adaptation and the effectiveness of a stimulus beyond its
onset. In R. H. Gilkey, & T. R. Anderson (Eds.), Binaural and spatial hearing in real and
virtual environments (pp. 211−232). Mahwah, NJ: Lawrence Erlbaum.
Hermans, D., & De Houwer, J. (1994). Affective and subjective familiarity ratings of 740
Dutch words. Psychologica Belgica, 34(2–3), 115−139.
Hopkins, W. D. (2008). Theoretical speculations on the evolutionary origins of
hemispheric specialization. Current Directions in Psychological Science, 17,
233−237.
Josse, G., & Tzourio-Mazoyer, N. (2004). Hemispheric specialization for language. Brain
Research Reviews, 44(1), 1−12.
Kensinger, E. A., & Corkin, S. (2003). Memory enhancement for emotional words: Are
emotional words more vividly remembered than neutral words? Memory and
Cognition, 31(8), 1169−1180.
Kinsbourne, M. (1970). The cerebral basis of lateral asymmetries in attention. Acta
Psychologica, 33, 193−201.
Koster, E. H. W., Crombez, G., Verschuere, B., & De Houwer, J. (2004). Selective attention
to threat in the dot probe paradigm: Differentiating vigilance and difficulty to
disengage. Behaviour Research and Therapy, 42(10), 1183−1192.
Kubovy, M. (1988). Should we resist the seductiveness of the space-time, visionaudition analogy. Journal of Experimental Psychology: Human Perception and
Performance, 14(2), 318−320.
Kubovy, M., & Van Valkenburg, D. (2001). Auditory and visual objects. Cognition, 80(1–2),
97−126.
Lavie, N. (1995). Perceptual load as a necessary condition for selective attention. Journal
of Experimental Psychology: Human Perception and Performance, 21(3), 451−468.
MacKay, D. G., & Ahmetzanov, M. V. (2005). Emotion, memory, and attention in the
taboo Stroop paradigm. An experimental analogue of flashbulb memories. Psychological Science, 16(1), 25−32.
MacKay, D. G., Hadley, C. B., & Schwartz, J. H. (2005). Relations between emotion,
illusory word perception, and orthographic repetition blindness: Tests of binding
theory. Quarterly Journal of Experimental Psychology: Human Experimental Psychology, 58(8), 1514−1533.
MacKay, D. G., Shafto, M., Taylor, J. K., Marian, D. E., Abrams, L., & Dyer, J. R. (2004).
Relations between emotion, memory, and attention: Evidence from taboo Stroop,
lexical decision, and immediate memory tasks. Memory and Cognition, 32(3),
474−488.
MacLeod, C., & Mathews, A. (1988). Anxiety and the allocation of attention to threat.
Quarterly Journal of Experimental Psychology, 40A(4), 653−670.
MacLeod, C., Mathews, A., & Tata, P. (1986). Attentional bias in emotional disorders.
Journal of Abnormal Psychology, 95(1), 15−20.
Mather, M., Mitchell, K. J., Raye, C. L., Novak, D. L., Greene, E. J., & Johnson, M. K. (2006).
Emotional arousal can impair feature binding in working memory. Journal of
Cognitive Neuroscience, 18(4), 614−625.
Mathewson, K. J., Arnell, K. M., & Mansfield, C. A. (2008). Capturing and holding
attention: The impact of emotional words in rapid serial visual presentation.
Memory and Cognition, 36(1), 182−200.
McKenna, F. P., & Sharma, D. (2004). Reversing the emotional Stroop effect reveals that
it is not what it seems: The role of fast and slow components. Journal of
Experimental Psychology: Learning, Memory, and Cognition, 30(2), 382−392.
Messina, D., Morais, J., & Cantraine, F. (1989). Valeur affective de 904 mots de la langue
française. Cahiers de Psychologie Cognitive, European Bulletin of Cognitive Psychology,
9(2), 165−187.
Mitchell, D. G. V., Nakic, M., Fridberg, D., Kamel, N., Pine, D. S., & Blair, R. J. R. (2007). The
impact of processing load on emotion. NeuroImage, 34, 1299−1309.
Mogg, K., & Bradley, B. P. (1998). A cognitive-motivational analysis of anxiety. Behaviour
Research and Therapy, 36, 809−848.
Mogg, K., & Bradley, B. P. (1999). Some methodological issues in assessing attentional
biases for threatening faces in anxiety: A replication study using a modified version
of the probe detection task. Behaviour Research and Therapy, 37, 595−604.
Mogg, K., Bradley, B. P., De Bono, J., & Painter, M. (1997). Time course of attentional bias
for threat information in non-clinical anxiety. Behaviour Research and Therapy, 35
(4), 297−303.
Mogg, K., Bradley, B. P., Dixon, C., Fisher, S., Twelftree, H., & McWilliams, A. (2000). Trait
anxiety, defensiveness and selective processing of threat: An investigation using
two measures of attentional bias. Personality and Individual Differences, 28,
1063−1077.
Mogg, K., Bradley, B. P., & Williams, R. (1995). Attentional bias in anxiety and
depression: The role of awareness. British Journal of Clinical Psychology, 34, 17−36.
Mogg, K., Holmes, A., Garner, M., & Bradley, B. P. (2008). Effects of threat cues on
attentional shifting, disengagement and response slowing in anxious individuals.
Behaviour Research and Therapy, 46, 656−667.
Most, S. B., Smith, S. D., Cooter, A. B., Levy, B. N., & Zald, D. H. (2007). The naked truth:
Positive, arousing distractors impair rapid target perception. Cognition and Emotion,
21(5), 964−981.
New, B., Pallier, C., Ferrand, L., & Matos, R. (2001). Une base de données lexicales du
français contemporain sur internet: LEXIQUE.L'Année Psychologique, 101, 447−462
http://www.lexique.org.
Nielsen, S. L., & Sarason, I. G. (1981). Emotion, personality, and selective attention.
Journal of Personality and Social Psychology, 41(5), 945−960.
Okon-Singer, H., Tzelgov, J., & Henik, A. (2007). Distinguishing between automaticity
and attention in the processing of emotionally significant stimuli. Emotion, 7(1),
147−157.
Pessoa, L., McKenna, M., Gutierrez, E., & Ungerleider, L. G. (2002). Neural processing of
emotional faces requires attention. Proceedings of the National Academy of Sciences,
99(17), 11458−11463.
Pessoa, L., Padmala, S., & Morland, T. (2005). Fate of unattended fearful faces in the
amygdala is determined by both attentional resources and cognitive modulation.
NeuroImage, 28, 249−255.
Posner, M. I. (1980). Orienting of attention. Quarterly Journal of Experimental Psychology,
32, 3−25.
Posner, M. I., & Cohen, Y. A. (1984). Components of Visual Orienting. In H. Bouma, & D. G.
Bouwhuis (Eds.), Attention and Performance X (pp. 531−554). Hillsdale, NJ:
Erlbaum.
Posner, M. I., Snyder, C. R. R., & Davidson, B. J. (1980). Attention and the detection of
signals. Journal of Experimental Psychology: General, 109, 160−174.
Ramon y Cajal, S. (1898). Estructura del kiasma optico y teoria general de los
entrecruzamientos de las vias nerviosas. Revista Trimestral Micrografica, 3, 15−65.
Rauschecker, J. P. (1997). Processing of complex sounds in the auditory cortex of cat,
monkey, and man. Acta Oto-Laryngologica — Supplement, 532, 34−38.
Rauschecker, J. P. (1998). Cortical processing of complex sounds. Current Opinions in
Neurobiology, 288, 516−521.
Rauschecker, J. P. (1998). Parallel processing in the auditory cortex of primates.
Audiology and Neuro-otology, 3, 86−103.
Rhodes, G. (1987). Auditory attention and the representation of spatial information.
Perception & Psychophysics, 42(1), 1−14.
Rozin, P., & Royzman, E. B. (2001). Negativity bias, negativity dominance, and
contagion. Personality and Social Psychology Review, 5, 296−320.
Russell, N. L., & Voyer, D. (2004). Reliability of laterality effects in a dichotic listening
task with words and syllables. Brain and Cognition, 54(3), 266−267.
Salemink, E., van den Hout, M. A., & Kindt, M. (2007). Selective attention and threat:
Quick orienting versus slow disengagement and two versions of the dot probe task.
Behaviour Research and Therapy, 45(3), 607−615.
Schimmack, U. (2005). Attentional interference effects of emotional pictures: Threat,
negativity or arousal? Emotion, 5(1), 55−66.
Please cite this article as: Bertels, J., et al., Emotional valence of spoken words influences the spatial orienting of attention, Acta Psychologica
(2010), doi:10.1016/j.actpsy.2010.02.008
ARTICLE IN PRESS
J. Bertels et al. / Acta Psychologica xxx (2010) xxx–xxx
Schmitt, M., Postma, A., & De Haan, E. (2000). Interactions between exogenous auditory
and visual spatial attention. The Quarterly Journal of Experimental Psychology, 53A
(1), 105−130.
Siegrist, M. (1995). Effects of taboo words on colour-naming performance on a Stroop
test. Perceptual and Motor Skills, 81(3), 1119−1122.
Smith Erthal, F., De Oliveira, L., Mocaiber, I., Garcia Pereira, M., Machado-Pinheiro, W.,
Volchan, E., et al. (2005). Load-dependent modulation of affective picture
processing. Cognitive, Affective, & Behavioral Neuroscience, 5(4), 388−395.
Spence, C. J., & Driver, J. (1994). Covert spatial orienting in audition: Exogenous and
endogenous mechanisms. Journal of Experimental Psychology: Human Perception
and Performance, 20(3), 555−574.
Spielberger, C. D. (1983). Manual for the State-Trait Anxiety Inventory (Form Y) (“Selfevaluation questionnaire”). Palo Alto, CA: Consulting Psychologists Press.
Stormark, K. M., Nordby, H., & Hugdahl, K. (1995). Attentional shifts to emotionally
charged cues: Behavioural and ERP data. Cognition and Emotion, 9(5), 507−523.
15
Suttie, I. D. (1926). A theory of decussation. Note on a possible adaptive significance of
the translateration of the upper motor and sensory neurones. The Journal of
Neurology and Psychopathology, VI, 267−280.
Taylor, S. E. (1991). Asymmetrical effects of positive and negative events: The
mobilization-minimization hypothesis. Psychological Bulletin, 110, 67−85.
Taylor, S. F., Kornblum, S., Lauber, E. J., Minoshima, S., & Koeppe, R. A. (1997). Isolation of
specific interference processing in the Stroop task: PET activation studies.
NeuroImage, 6(2), 81−92.
Thomas, L. A., & LaBar, K. S. (2005). Emotional arousal enhances word repetition
priming. Cognition and Emotion, 19(7), 107−1047.
Van der Goten, K., De Vooght, G., & Kemps, E. (1999). Concreteness and emotional
valence ratings of 399 Dutch nouns. Psychologica Belgica, 39(1), 49−70.
Zajonc, R. B. (1962). Response suppression in perceptual defense. Journal of
Experimental Psychology, 64(3), 206−214.
Please cite this article as: Bertels, J., et al., Emotional valence of spoken words influences the spatial orienting of attention, Acta Psychologica
(2010), doi:10.1016/j.actpsy.2010.02.008