Positive vibes, predictive minds: positive mood fuels cross-modal semantic predictions in high-constraint contexts

Semantic processing involves prediction. Recent theoretical frameworks suggest that prediction can be shaped by mood, our background affective state. Here we show that positive mood dynamically promotes cross-modal semantic predictions in richer semantic contexts. Participants were induced into positive and negative moods. They saw three-picture sequences varying in semantic constraint, followed by semantically related or unrelated words. First, context construction (indexed by N400 responses) was enhanced in medium-constraint condition under positive compared to negative mood. Second, anticipatory pre-activation (indexed by SPN responses) was facilitated under positive compared to negative mood in high-constraint context. Finally, re-evaluation of prediction matches (indexed by LFP responses) was facilitated in high-constraint context under positive compared to negative mood. Together, these findings indicate that positive mood shapes neurocognitive mechanisms underlying cross-modal semantic predictions in semantically richer contexts, facilitating context construction and prediction generation, anticipatory pre-activation of linguistic representations, and integration of confirmed predictions.