Effects of emotional spoken words on exogenous attentional orienting

Abstract

Attentional biases linked to emotional stimuli were investigated in healthy people using an auditory adaptation of the cueing paradigm. Specifically, we investigated whether both validity effects elicited by predictive, endogenous cues and the Inhibition of Return phenomenon (IOR; Posner & Cohen, 1984) elicited by unpredictive, exogenous cues are influenced by the emotional content of spoken words. Supporting the idea that exogenous orienting is not an encapsulated phenomenon (Stolz, 1996), we found abolished IOR for negative words (Experiments 3 and 4). Thus, attention would not be prevented from returning to the previously explored location of a negative word. On the contrary, no emotional modulation of the validity effects was observed (Experiments 1 and 2), suggesting that the intervention of resource-demanding orienting strategies increased cognitive load and thus prevented any emotional modulation. Still, facilitative, nonspatial effects of negative words were found when initial attentional shifts elicited by the cue were both exogenous and endogenous (Experiment 1), but not when they were exclusively endogenous (Experiment 2). These results highlight the importance of both the negativity of a stimulus and the automaticity of attentional shifts in eliciting spatial and nonspatial attentional effects.

Keywords:

• Cueing
• Emotional valence
• Inhibition of return
• Validity effects

Acknowledgements

JB was Research Fellow of the Fonds de la Recherche Scientifique (FRS-FNRS) and is presently Scientific Research Worker of the FRS-FNRS. RK is Senior Research Associate of the FRS-FNRS. All authors were affiliated to the Research Unit in Cognitive Neurosciences (UNESCOG), Université Libre de Bruxelles. This work was also supported by an FRS-FNRS grant (1.5705.06).

Notes

¹For sake of homogeneity, a neutral instead of an emotional tone of voice was used in the present study (see Bertels et al., 2010, in press). Indeed, we wanted first to compare emotional and neutral words according to one dimension (i.e., their emotional content), without potentially confounding this effect with the effect of another dimension, namely the emotional prosody (see Bertels et al., 2009, for data illustrating the importance of the latter).

²When trait-and state-anxiety T scores as covariates, this interaction remained significant, F(3, 54)=7.048, MSE=1277.1, p<.001.

³Unfortunately, the anxiety questionnaires administered to the participants of Experiment 4 were lost. We are thus unable to provide the anxiety scores relative to these subjects as well as to include these data in a correlation analysis.

⁴Coherently, comparing across experiments the attentional effects obtained for each word type revealed an interaction between word type and experiment that approached the conventional significance level, F(9, 246)=1.839, MSE=852.917, p=.062. This interaction reflects the fact that the effect of word type was significant in Experiments 3 and 4 (p<.05 and p=.01, respectively), but not in Experiments 1 and 2 (both ps>.10).

⁵Similarly, the participant's gender should be balanced in each experiment, since emotional stimuli were found to be processed in a different way as a function of gender (e.g., Egloff & Schmukle, 2004). In addition, further studies should also control the gender of the voice pronouncing the word cues. However, participant's gender is rarely taken into account in studies investigating emotional influences on attention, and most of these studies actually use unbalanced samples of male and female participants. Moreover, in a previous study (Bertels et al., in press), controlling this variable revealed that the participant's gender did not interact with the effect of emotional words on the nonspatial orienting of attention.