Cognitively demanding stimuli can acquire positive valence.

Generally, people tend to avoid stimuli that require mental effort; effort can generate negative emotions. However, employing mental effort can also promote positive emotions, given a successful outcome. We investigated whether the level of cognitive effort associated with stimuli will elicit positive or negative emotions. In Experiment 1, participants performed a gender Stroop task during the association phase. The actors from the Stroop task expressed emotionlessness, while half of the actors were displayed in the mostly incongruent (MI) condition and the rest in the mostly congruent (MC) condition. In the transfer phase, we used the same actors for the emotion discrimination task, and the actors expressed a positive emotion half of the time and a negative emotion for the other half. For the MI actors, participants responded faster to positive emotion than to negative emotion, but this difference was not significant for the MC actors. In Experiment 2, the association phase involved a task switching paradigm in which half of the actors were presented in the mostly switching (MS) condition and the other in the mostly repetition (MR) condition. In the transfer phase, the same individuals' faces were used for emotion discrimination. For the MS actors, but not the MR actors, the responses were faster to positive emotion than to negative emotion. Our results imply that stimuli associated with more cognitive effort (i.e., MI and MS stimuli) may be perceived as more positive after a successful outcome of a task, although future research is required to replicate these findings.

Stimuli with a positive valence can facilitate cognitive control.

In the process of interacting with people and objects, humans assign affective valence. By using an association-transfer paradigm, the current study investigated whether the emotion associated with a stimulus would have an impact on cognitive control outcomes. During the association phase of two experiments reported here, participants identified the emotion expressed by an actor's face as either positive (i.e., smiling) or negative (i.e., frowning). Half of the actors expressed positive emotions (MP) on 80% of trials, while the other half expressed negative emotions (MN) on 80% of trials. We tested the cognitive effect of these associations in two experiments. In the transfer phase of Experiment 1, the same actors from the association phase were shown with neutral expression during a gender Stroop task, requiring participants to identify the gender of the face while ignoring a gender word (congruent or incongruent) that was imposed upon the face. The Stroop effect was significant for the MN faces, but the effect disappeared for the MP faces. In the transfer phase of Experiment 2, the emotionless faces were presented in a task-switching paradigm, in which participants identified the age (i.e., old or young) or the gender depending on the task cue. The task switch cost was smaller (though significant) for the MP faces than for the MN faces. These results suggest that, relative to social stimuli associated with negative expressions, social stimuli associated with positive expressions can promote better cognitive control and inhibit distractor interference in goal-oriented behavior.

Subcortical regions of the human visual system do not process faces holistically.

Face perception is considered to be evolutionarily adaptive and conserved across species. While subcortical visual brain areas are implicated in face perception based on existing evidence from phylogenetic and ontogenetic studies, whether these subcortical structures contribute to more complex visual computations such as the holistic processing (HP) of faces in humans is unknown. To address this issue, we used a well-established marker of HP, the composite face effect (CFE), with a group of adult human observers, and presented two sequential faces in a trial monocularly or interocularly using a Wheatstone stereoscope. HP refers to the finding that two identical top (or bottom) halves of a face are judged to be different when their task-irrelevant bottom (or top) halves belong to different faces. Because humans process faces holistically, they are unable to ignore the information from the irrelevant half of the composite face, and this is true to an even greater extent when the two halves of the faces are aligned compared with when they are misaligned ('Alignment effect'). The results revealed the HP effect and also uncovered the Alignment effect, a key marker of the CFE. The findings also indicated a monocular advantage, replicating the known subcortical contribution to face perception. There was, however, no statistically significant difference in the CFE when the images were presented in the monocular versus interocular conditions. These findings indicate that HP is not necessarily mediated by the subcortical visual pathway, and suggest that further investigation of cortical, rather than subcortical, structures might advance our understanding of HP and its role in face processing.

Perceived conflict may be negative but resolved conflict is not.

The current study investigated how exposure to a conflict stimulus influences the judgment of a subsequent stimulus's valence. We used an affective priming paradigm, presenting a color Stroop stimulus as a prime and a face as a target for an emotion recognition task. When the task for the prime was passive viewing (Experiment 1), congruent primes resulted in faster responses to emotionally positive targets than negative targets. However, this positivity bias disappeared following incongruent primes. In Experiment 2, instead of passive viewing, participants were asked to indicate the congruency of the prime, and the positivity bias was significant following the congruent prime but not following the incongruent prime. In Experiments 3 and 4, participants performed the conventional Stroop task on the prime, therefore resolving the conflict when the prime was incongruent. Experiment 3 adopted an equal proportion of congruent and incongruent primes. Experiment 4 adopted twice as many congruent primes as incongruent primes. In both experiments, the positivity bias was not significant regardless of the congruency of the prime. These results suggest that detecting conflict may interfere with positive affect or promote negative affect, therefore reducing the positivity bias. Once the conflict is resolved, however, the negative valence may disappear.