Emotional Background Music
Gwendoline Oslager
We examined the moderating influence of dispositional behavioral inhibition system (BIS) and behavioral activation system (BAS) sensitivities on the relationship of startling background music with emotion-related subjective and physiological responses elicited during reading news reports, and with memory performance among 26 adult men and women. Physiological parameters measured were respiratory sinus arrhythmia (RSA), electrodermal activity (EDA), and facial electromyography (EMG). The results showed that, among high BAS individuals, news stories with startling background music were rated as more interesting and elicited higher zygomatic EMG activity and RSA than news stories with non-startling music. Among low BAS individuals, news stories with startling background music were rated as less pleasant and more arousing and prompted higher EDA. No BIS-related effects or effects on memory were found. Startling background music may have adverse (e.g., negative arousal) or beneficial effects (e.g., a positive emotional state and stronger positive engagement) depending on dispositional BAS sensitivity of an individual. Actual or potential applications of this research include the personalization of media presentations when using modern media and communications technologies.
Purchase intention was stronger in the group that watched the ad with background music. In the seven-point scale questionnaire, the average purchase intention of the group who heard background music was a 4.36, while the group without music averaged 3.64.
Another element that could explain the variability between the results of previous studies is the lack of control over the emotional characteristics of the sound stimuli being utilized (as discussed in Kämpfe et al., 2010; Schellenberg, 2013). Indeed, different sound environments can induce different emotions. Particularly for musical stimuli, musical parameters, such as tempo, can be modulated to induce different musical emotions, like the level of arousal (Gabrielsson and Juslin, 2003); music with fast tempi are usually considered as stimulating, while music with slow tempi are considered as relaxing (Västjäll, 2001; Bigand et al., 2005; Vieillard et al., 2008). The emotional characteristics of a sound stimuli, like its level of arousal, are important to consider as studies have shown links between them and performance on cognitive tasks (e.g., spatial skills, Thompson et al., 2001; selective attention, Ghimire et al., 2019). Indeed, according to the arousal-mood hypothesis (Nantais and Schellenberg, 1999; Thompson et al., 2001, 2011; Husain et al., 2002; Schellenberg and Hallam, 2005; Schellenberg et al., 2008), cognitive performance can be promoted by sound stimulation, notably by increasing physiological activation and improving mood. Both music and noise can induce emotions (Hunter and Schellenberg, 2010), but there is a general agreement that music is efficient to induce positive emotions, and therefore it can be employed to positively modulate mood (Thompson et al., 2001). The previously cited research by Schellenberg and Weiss (2013) has shown that when participants listen to music that positively alters their mood before performing a cognitive task, like a stimulating and pleasant music, their performance in this cognitive task was improved. The arousal-mood hypothesis has been built on data that are based on listening to a stimulus before the accomplishment of a cognitive task.
These results are somewhat encouraging as they showed that the addition of to-be processed cognitive information (e.g., background music/noise) does not necessarily have deleterious effects on attentional performance as some theories suggests (e.g., Kahneman (1973) limited capacity model). With these results, it is possible to assume that performing a task requiring attention in the presence of instrumental music should not have a negative effect on the level of selective attention demand in order to perform the task optimally.
The arousal-mood hypothesis (Nantais and Schellenberg, 1999; Thompson et al., 2001, 2011; Husain et al., 2002; Schellenberg and Hallam, 2005; Schellenberg et al., 2008) suggests that a sound environment judged to be stimulating and pleasant would be a beneficial environment to optimize cognitive performance (for details, see Schellenberg, 2013). It was therefore expected that the stimulating music condition would be the sound environment in which we would see the lowest error rate and weakest Stroop interference. In contrast to the hypotheses, the presence of pleasant and stimulating music during the accomplishment of the task did not significantly improve task performance. A small tendency to make more errors on incongruent trials in this sound environment was also noted. These results differ from those of previous work studying the effect of the arousal level of background music upon selective attention (Fernandez et al., 2019; Cloutier et al., 2020). However, these studies mainly aimed to make comparisons between groups (elderly vs. young adults), while the present study had an objective of generalization to the adult population. Furthermore, the tasks involved were different: while previous studies employed the Flanker task to assess selective visual attention, the current study utilized the Stroop task which involves language processing. On the other hand, the number of sound conditions in this study may affect the statistical power of the results. It would therefore be interesting to investigate whether the results would be the same with even a larger sample-size in future studies (even though our sample-size was larger than in previous studies).
A key finding of this study is a negative effect of music-matched noise stimuli (low pleasantness) on attentional performance. These results converge with previous work by Masataka and Perlovsky (2013) and Slevc et al. (2013) showing lower performance on a similar Stroop task in the presence of dissonant music (sound pairings perceived as generally unpleasant or possessing low-pleasantness valence). Interestingly in these studies, greater consonance (sound pairings perceived as generally pleasant or possessing a high-pleasantness valence) led to better performance on the Stroop task. It would therefore be interesting to investigate further to assess which factor, the level of valence/pleasantness or the degree of consonance, had the greater influence upon the results of this and previous studies (Masataka and Perlovsky, 2013; Slevc et al., 2013). This could be done by integrating stimuli that are both consonant and unpleasant, such as scary or sad music, or by specially composed music material. In previous research, the relationship between background music and cognitive performance seems to be affected by the degree of familiarity of the musical stimulus (if the music was already known to the participant). Higher familiarity has a positive effect on performance for cognitive tasks (Darrow et al., 2006; Speer, 2011; Giannouli, 2012). One potential limitation of this study is that, despite an attempt to select equally familiar music of similar valence, the stimulating musical stimuli were rated as more familiar by the participants than the relaxing musical stimuli (see Supplementary Materials for details). It is then surprising that the present findings did not support an effect of stimulating music on task performance given that the stimulating music condition was biased toward higher familiarity.
Judgments of valence can be influenced by the familiarity of a musical piece. Some studies have shown that perceivers tend to find a stimulus that they already know (e.g., a piece of music) more pleasant (Parente, 1976; Schellenberg et al., 2008; Van Den Bosch et al., 2013). Familiar background music has also been associated with increased pleasure in the process of completing a task without compromising task performance (Pereira et al., 2011; Feng and Bidelman, 2015). In this regard, Darrow et al. (2006); Speer (2011), Giannouli (2012), and Kiss and Linnell (2020) asked their participants to bring their favorite music into the lab, which then was used as background music to perform a selective attention task or a sustained-attention task. In these studies, the music selected by the participants held characteristics of high emotional valence and familiarity. However, the other characteristics of the music utilized were heterogeneous between participants (e.g., style, complexity of the music pieces, presence of lyrics, or the level of arousal). The results of these studies indicate that participants consistently performed better in the familiar music conditions. As we know little about the characteristics of the different pieces of music used in these studies and that a great variability is present between them, it is difficult to identify whether the results are generalizable to listening to background music in general or whether they are specifically attributable to a modulation of mood and/or arousal due to the emotional characteristics and familiarity of the music used. Future research should combine this approach with systematic acoustic as well as musical and linguistic structure analyses of the used material to further our understanding of the potential characteristics involved in the observed effects.
Taken together, our findings suggest that it is not sufficient for background music to be arousing, pleasant and familiar in order to enhance attentional performance as suggested by the Arousal-mood theory, and that factors related to individual musical taste may be driving the effects found in previous studies.
Just as music can alter expectations and interpretations of a visual scene, thecausal connection between music and cognition (e.g., evaluations) of a situationappears to work both ways, in that the perceived emotional tone of the accompanyingmusic can be influenced by visual aspects of the scene or situation, such as thefacial expression of the performer (Thompson,Graham, & Russo, 2005). Although there is a large body of literatureon the topic of encoding specificity (Tulving &Thompson, 1973), and the presence versus absence of encoding cues atretrieval (Tulving & Osler, 1968; Tulving & Pearlstone, 1966), little workappears to have taken place examining the effects of the emotional tone ofbackground music on memory for visual emotional stimuli. This connection forms thefocus of the work presented here. Specifically, we were interested in knowingwhether or not the presence of emotion-conveying background music while studyingpictures would influencesubsequent memory for the facial expressions of charactersdepicted in those pictures. Generally, the perception of facial expression involvesa number of diverse and interactive brain regions that make use of both the featuresin the face itself and of any auxiliary emotion-conveying contextual information(Adolphs, 2006) in arriving at a perceivedemotional tone being conveyed in a scene. If it is the case that any aspect ofcontext could potentially contribute to the perception of a facial expression, thenthe emotional tone set by the music paired with a picture should influence recall ofemotional aspects of the picture when subsequently tested, by biasing responsestoward the mood conveyed by the music. That is, if during encoding a photo with anindividual displaying a particular facial expression (e.g., smiling) appears inconjunction with a musical excerpt that conveys conflicting emotional information(sadness), will the subsequent recall of the emotional facial expression beinfluenced by the emotional musical context during encoding?