Because emotional stimuli are motivationally relevant, they capture attention. This idea is common to theories such as motivated attention, emotional attention, and natural selective attention. According to the motivational model of emotion
[1, 4], emotional stimuli such as predators, food, or sexual scenes activate either the defensive or appetitive motivational system, thus preparing the organism for appropriate action. Which system (defensive or appetitive) is activated corresponds to the subjective experience of valence, and the degree of activation corresponds to the subjective experience of arousal
Allocation of attention to emotional pictures has been studied by means of event-related potentials (ERPs). Two commonly used measures are the early posterior negativity (EPN) and the late positive potential (LPP). Both reflect the stronger allocation of attentional resources to emotional than neutral pictures
[5, 6]. When the ERP waves to neutral pictures are subtracted from the ERP waves to emotional pictures (e.g., mutilation or erotica), the early posterior negativity (EPN) is a negative deflection starting about 200 ms after picture onset and is evident over electrodes in the temporal-occipital region
. Similarly, the ERP waves to the pictures reveal a late positive potential (LPP) that is seen over parietal-central electrodes from about 300 ms after stimulus onset and that is larger for emotional than neutral pictures
. Both the EPN and the LPP can be observed without explicit instruction to attend to the emotional content
. Studies on the neural mechanisms have found correlations between LPP amplitudes to emotional pictures (of negative or positive valence) and fMRI activations in the lateral occipital, inferotemporal, and parietal visual areas
 and in the insula, anterior cingulate, ventral striatum/nucleus accumbens, and amygdala
. Also, a recent combined EEG/fMRI study suggests that these different areas contribute differently to LPP amplitudes depending on valence
. In sum, these findings for the EPN and LPP to emotional (vs. neutral) pictures show that emotional stimuli capture attention without explicit instruction and thus, are intrinsically motivational.
The stronger allocation of attentional resources to emotional than neutral pictures, as indexed by the EPN and LPP, has also been examined in phobia. Phobia is the most common anxiety disorder, with a lifetime prevalence ranging from 10% to 18%
[12, 13]. The most common phobia is fear of spiders, with a prevalence of 1.2% for men and 5.6% for women
. Phobia is usually thought to be the result of aversive conditioning
, that is, Pavlovian or classical conditioning where a biologically innate fear response is coupled with a stimulus that does not by itself elicit it. In support, brain imaging show similar activation patterns for phobia and experimentally induced aversive conditioning in humans
ERP studies have confirmed that feared pictures capture attention. In people who are diagnosed with spider phobia or who report high levels of spider fear, the LPP amplitudes are enhanced to spiders versus neutral pictures
[17–26], as are EPN amplitudes
These ERP findings of larger EPN and LPP amplitudes to spiders than to neutral pictures in participants with high levels of spider fear (or phobia) are relevant to theories that argue that attentional biases to threat are important factors in the development of anxiety disorders
[28, 29]. However, it is unclear whether the ERP findings mean that in people with spider fear, attention to spider pictures exceeds the level of attention expected from the spider pictures’ high arousal and negative valence (i.e., their intrinsic motivational value). To study the effect of spider pictures on attention over and above the level of attention that is expected from the pictures’ high arousal and negative valence, it is necessary to compare spider pictures to fear-irrelevant, negative pictures in people with spider fear
. Specifically, if people with spider fear rate spiders and other negative pictures similarly in terms of valence and arousal but attend more strongly to the spiders than to the other negative pictures, then this finding would demonstrate that attention to spider pictures in people with spider fear exceeds the level of attention expected from the spider pictures’ high arousal and negative valence. Unfortunately, for the LPP, studies did not analyze arousal and valence ratings across picture types
[23, 26], or studies did not match picture types in these regards
[19, 22]. For the EPN, studies did not report whether emotion ratings differed between the picture types
. Also, another study on the EPN compared spider pictures that had mainly simple picture composition (i.e., figures) with other negative pictures that had mainly complex composition (i.e., scenes)
. However, this design confounds emotion with picture composition because the EPN is smaller to emotional scenes than to emotional figures even if valence and arousal are matched
. Taken together, although there is strong evidence that emotional pictures draw attention, as indexed by the EPN and the LPP, it is unresolved whether the effect on attention from spider pictures in people with spider fear exceeds the level that would be expected on the basis of the spider pictures’ intrinsic motivational relevance (i.e., high arousal and negative valence).
To study whether emotional pictures capture attention even if the pictures are task irrelevant (i.e., distracters), many studies have investigated emotional responses during spatial inattention, that is, by comparing responses to pictures in attended with those in unattended locations
[31, 32]. The relevant ERP studies on this issue have mainly used emotional pictures from the International Affective Picture System (IAPS)
. Results showed that LPP amplitudes to negative and positive IAPS pictures and pictures of emotional faces were eliminated when the pictures were presented in the periphery and attention was directed at fixation
[34–37], and strongly reduced if not eliminated when pictures were presented at fixation and attention was directed to the periphery
[30, 38–43]. Similar results were reported for studies on the EPN with reductions of amplitudes when emotional pictures were presented in unattended locations
[30, 34], even though some studies reported null findings
[40–42]. However, it has yet to be studied whether spatial inattention decreases the EPN and the LPP to spiders in spider fear. Also, it is unresolved whether effects of spatial inattention decrease the EPN and the LPP less strongly for spiders than for fear-irrelevant negative pictures in spider fear. If so, this finding would suggest that attention is drawn to the feared spiders per se over and above their intrinsic motivational value (i.e., high arousal, negative valence).
Effects of distracting emotional pictures on attention can also be studied by presenting task-irrelevant stimuli while participants perform a concurrent task that varies in attentional demands. According to Load theory
[44, 45], attention is a resource that can be distributed across tasks. If a task is taxing only a small amount of this resource (low perceptual load), there are spare resources that are used to process task-irrelevant stimuli. However, if a task is consuming all resources (high perceptual load), there are no resources left and consequently no task-irrelevant stimuli will be processed. Notably, a recent study with spider fear suggested that Load theory may not apply fully to spider pictures in spider fear
. In this study, participants with and without high levels of spider fear were shown task-irrelevant pictures of spiders and mushrooms at fixation while participants performed a discrimination task on letters that surrounded the pictures with either three letters (low load) or six letters (high load). Results showed that enhanced LPP amplitudes to spider pictures in fearful participants did not differ between low and high load (with negligible effect size). Thus, perceptual load did not seem to influence processing of spiders in spider fear.
However, several subsequent studies reported similar null effects for generally negative pictures from the IAPS picture system (e.g., mutilation, disgust)
[40, 41]. Participants performed a letter discrimination task while pictures were shown at fixation. Performance on the letter discrimination task decreased strongly with load. Although participants showed generally enhanced EPN and LPP amplitudes to negative pictures, the EPN and LPP amplitudes did not differ between low and high load. Similarly, when participants performed either a detection task (low load) or a discrimination task (high load) on symbols at fixation, the steady-state visual evoked potentials (ssVEP) to flickering negative IAPS pictures in the background were unaffected by load
. In contrast to these null findings, one study found that the EPN to IAPS pictures (positive and negative valence) versus neutral pictures decreased when task demands increased
. In the study, lines were superimposed on the pictures, and participants had to count the number of line trials. Task demands were increased by increasing the proportion of line trials. If it is argued that this manipulation increases perceptual load, these findings suggest that perceptual load reduces EPN to generally emotional pictures. However, spiders and other negative pictures in participants with high spider fear have not been included in the same study. Therefore, it is possible that in spider fear, effects of perceptual load decrease the EPN and the LPP less strongly for spiders than for fear-irrelevant negative pictures in spider fear, thus suggesting that attention is drawn to the feared spiders per se.
To summarize, theories on the etiology of anxiety disorders assign an important causal role to attentional biases
[28, 29]. Previous studies confirm that people with high spider fear attend more strongly to spiders than to neutral pictures, as indexed by enhanced EPN and LPP amplitudes to spiders versus neutral pictures. Further, previous research has studied effects of manipulations of spatial attention and perceptual load on EPN and LPP amplitudes for generally negative pictures. However, no previous study has studied whether effects of manipulations of spatial attention and perceptual load differ for spiders and other negative pictures in people with high spider fear. If the EPN and LPP amplitudes are reduced less strongly for spiders than for other negative pictures in people with high levels of spider fear, then this finding would provide evidence that in spider fear, spiders have an effect on attention over and above their intrinsic motivational value (i.e., high arousal and negative valence). To address this question, the present study included participants with and without spider fear and recorded ERPs to spiders, negative pictures, and neutral pictures during manipulations of spatial attention and perceptual load. Attention to the pictures was indexed by the EPN and LPP. Because the EPN and LPP are affected by picture composition
[30, 48, 49], the different picture categories (spiders, negative pictures, and neutral pictures) were matched in picture composition to avoid any confounding effects from picture composition.