When studying for a test, an individual may incidentally receive an unrelated message providing bad news. Although this event is incidental and unrelated to the individual’s current goals, growing evidence from the literature indicates that incidental events can influence our emotions and behaviors without our intentional effort (Loewenstein and Lerner, 2003; Berkman and Lieberman, 2009). Indeed, it is well documented that emotional processing is extremely influenced by incidental events as different types of contextual factors can modulate behavior, electrocortical activity, and physiological and neural responses during the viewing of emotional pictures (Foti and Hajcak, 2008; Mocaiber et al., 2009, 2010, 2011; Oliveira et al., 2009). For instance, antecedent descriptions stating that pictures of mutilated bodies presented in an experiment were taken from movie productions can attenuate the emotional impact of unpleasant pictures (Mocaiber et al., 2009, 2010, 2011; Oliveira et al., 2009). In addition, some studies have found that incidental negative emotions or affective states can pervasively carry over from one event to next, affecting fairness decisions (Liu et al., 2016), cognitive capacity (Figueira et al., 2017) and emotion recognition (Qiao-Tasserit et al., 2017).

Importantly, the ability to identify emotional expressions can be considered an adaptive social behavior as emotional perception can influence the production and regulation of affective states (Bourke et al., 2010). Accumulating evidence from the literature indicates that facial stimuli may be subject to influences from contextual factors (de Gelder et al., 2006; Aviezer et al., 2008, 2011; Barrett et al., 2011; Wieser and Brosch, 2012), probably because facial expressions are typically embedded in a rich context in real life (Aviezer et al., 2011). Furthermore, facial expressions are extremely important for social interaction as they provide information about other people’s emotions and social intentions (Wieser and Brosch, 2012). It is also worth noting that different types of contexts are able to influence the perception and evaluation of faces (Kim et al., 2004; Aviezer et al., 2008; Righart and De Gelder, 2008).

Contextual information is particularly powerful when facial expressions are either ambiguous or unrelated to any specific emotion, such as neutral faces (Ekman and Friesen, 1976; Lee et al., 2008). Other recent evidence showing that contextual cues can influence our evaluation of neutral faces was obtained from studies using an original behavioral paradigm based on a version of the filmic “Kuleshov effect” (e.g., Pudovkin, 1970; Kuleshov, 1974; Barratt et al., 2016; Calbi et al., 2017, 2019). Furthermore, recent studies have demonstrated that the social context in which a neutral face is presented can impact participants’ emotional responses to such faces (Davis et al., 2010; Schwarz et al., 2013; Wieser et al., 2014; Klein et al., 2015; McCrackin and Itier, 2018). Such contextual modulation of neutral faces has been observed in studies using behavioral paradigms and electrocortical and neuroimaging studies (Davis et al., 2010; Schwarz et al., 2013; Suess et al., 2015).

For example, in a social conditioning study, Davis et al. (2010) demonstrated that neutral faces conditioned with positive and negative context cues subsequently provoked higher activations in the lateral ventral amygdala. Studying the contextual modulation of neutral faces can be especially important because the response to these faces can help to identify mental disorders (Leppänen et al., 2004; Jiao et al., 2011; Oliveira et al., 2013; Wieser and Moscovitch, 2015; Liu et al., 2016; Manelis et al., 2019).

However, little research has investigated whether emotional contextual information presented during a preceding and unrelated experiment could pervasively carry over from one experiment to the next to modulate neutral face processing. Notably, Pichon et al. (2012) demonstrated that prior exposure to negative words led to increased activation in the basal amygdala in a subsequent, unrelated visual task. It remains unknown, however, whether negative contextual information presented during a preceding and unrelated experiment could modulate the emotional judgment of neutral faces in a subsequent experiment. In the current study, we assessed whether an emotional text presented in a first experiment could attribute negative emotion to neutral faces in a second experiment. Specifically, our sample comprised two groups: in one group, participants read a neutral text about the loss of an identity card, and in the other group, participants read a negative text about sexual abuse. Subsequently, the participants performed a second experiment with two tasks unrelated to the previous task. The attentional task consisted of judging the orientation of two peripheral bars while pictures (neutral faces or objects) were presented as distractors, and the emotional judgment task consisted of judging the valence of the pictures (neutral faces or objects) while bars were presented as distractors. We predicted that negative text presented previously in the first experiment would modulate both subsequent tasks in the second experiment when compared to neutral text. Therefore, it was expected that the present study could contribute to elucidating the role of long-term modulation of incidental cues in our behavior and actions. This study is important because it highlights that incidental information may have the potential to implicitly change our judgments and emotions, directly affecting our everyday lives.

The main goal of the present study was to assess whether the emotional text presented in the first experiment could generate a negative bias in the processing of faces presented in the second and unrelated experiment using the following two tasks: (1) an attentional task in which neutral faces were presented as distractors and (2) a neutral face emotional judgment task. As expected, we found that negative text was judged as more unpleasant and arousing than neutral text, showing that our manipulation to induce a negative emotion in the first experiment was successful. Importantly, our results showed that the judgment index in the second experiment after reading negative text was significantly higher than that obtained after reading neutral text, indicating that participants rated more faces as negative in the negative context than in the neutral context. To the best of our knowledge, this is the first study to demonstrate that emotion information available in a previous experiment that was unrelated to the current goal could increase participants’ propensity to interpret neutral faces in a more negative way. However, one unexpected aspect of our findings is that for the bar judgment task, the reaction time index obtained in the negative context was not significantly different from that obtained in the neutral context. Therefore, our results further demonstrate that the attribution of emotion to neutral faces occurred only in the emotional judgment task.

Our findings are in line with previous literature indicating that neutral face stimuli can be subject to influences from contextual factors (Anderson et al., 2011; Suess et al., 2015; Wieser and Moscovitch, 2015; Zhao et al., 2017; Baum and Abdel Rahman, 2021). Moreover, these results add to the literature by showing that negative contextual information presented in a preceding and unrelated experiment could attribute emotion to neutral faces in a subsequent experiment with ∼12 min of duration and an interval of a few minutes (approximately 5 min) between the experiments. Our findings counter the idea that emotional context effects dissipate shortly after exposition, as in paradigms where emotional primes (e.g., words) are presented for durations shorter than a few hundred milliseconds (Hermans et al., 2001). Consistent with our findings, Qiao-Tasserit et al. (2017) show that a sustained transient emotional state induced by negative emotional clips outlasts the exposure period by ∼2 min and subsequently biases the interpretation of morphed facial expressions ranging from fear to happiness in a forced-choice emotion classification task. In fact, induced negative states increase participants’ propensity to modulate the interpretation of ambiguous expressions relative to induced neutral states. We further extend previous work by demonstrating that (1) we can attribute negative emotion that pervasively carries over from one experiment to the next, (2) the effect in the emotional judgment task is observed between groups, and (3) negative text generates incidental negative emotions that outlast the exposure period by at least ∼8 min (when the emotional judgment task was present first) and a maximum of 15 min.

In line with the idea that emotional context can have long-term effects after exposition, Pichon et al. (2012) found that prior exposure to negative words (relative to neutral words) modifies attention control in a subsequent, unrelated visual task. Our work differs from Pichon et al. (2012) as instead of asking the participants to match (same/different judgment) the stimulus pair at attended locations in an attention task, we used an emotional judgment in which the participants indicated whether the neutral face stimuli valence was neutral or unpleasant.

Notably, only females were chosen as participants in this study. This choice was made because some studies in the literature suggest that an interaction exists between gender and emotional perceptions of faces. For example, some studies have shown that women are better than men in distinguishing facial expressions (Thayer and Johnson, 2000; Guillem and Mograss, 2005). An interaction effect on memory, recognition, and judgment bias also exists between the gender of the face observed and the gender of the participants (Herlitz and Lovén, 2013; Wang, 2013; Harris et al., 2016). For instance, Hofmann et al. (2006) and Hoffmann et al. (2010) showed that women were more accurate than men in recognizing subtle facial displays of emotion. Another important issue is that we used text discussing sexual abuse in this study. In general, females are more likely to be victims of sexual violence than males (de Waal et al., 2017). Therefore, it is expected that the impact of this type of text is greater among females than males, even among females who have not suffered from this traumatic event.

We found an unexpected but interesting result for the bar orientation task: the reaction time index obtained in the negative context was not significantly different from that obtained in the neutral context. Similar to our study, Qiao-Tasserit et al. (2017) found that induced negative states bias the interpretation of facial expressions in perceptual decision-making tasks, but the reaction time measurements for this paradigm lacked the sensitivity necessary to detect this effect. An explanation may be that the reaction time is a more implicit measure compared to the emotional judgment of a neutral face for directly accessing the attribution of emotion generated by the preceding text. Therefore, cognitively assessed affective content can be more easily modulated by incidental information than less cognitive emotional content, such as that measured by reaction time. This notion is consistent with previous studies showing that participants rated neutral faces as negative significantly more in negative contexts (fearful and sadness) than in other contexts (neutral or happy/desire contexts) (Barratt et al., 2016; Calbi et al., 2017). However, more implicit measures, such as eye tracking, did not support the expected differences between the context conditions (Barratt et al., 2016). Furthermore, Deveney and Pizzagalli (2008) showed that their participants regulated their emotions to unpleasant pictures regardless of whether the word was negative or neutral, whereas their reaction time data failed to show emotion regulation modulation.

There were also some limitations in the present study. The main limitation was that after the experiment, we did not directly assess whether any participant guessed that the experiments were linked.

In summary, the present work adds to the literature important evidence that our behaviors and actions are modulated by preceding information in an incidental way. Previous negative information generated by reading a text was able to attribute emotion to neutral faces in a subsequent experiment presented several minutes later. These results promote reflection on how much deliberate and conscious control we have over our decisions, actions, and attitudes in our everyday life.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

The studies involving human participants were reviewed and approved by the Research Ethics Committee of Fluminense Federal University. The patients/participants provided their written informed consent to participate in this study.

MN analyzed and interpreted the data, wrote the manuscript, and contributed to the data organization. LP contributed to the data organization, data interpretation, and revising the manuscript. IM, ID, and BR interpreted the data and revised the manuscript. MP contributed to writing and revising the manuscript and the data interpretation. LO analyzed, interpreted the data, wrote, and revised the manuscript. All authors read and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.