EXPRESS: Examining the Memory Effects of Emotional Ambiguity via the Lens of Emotional Context Variability

The emotional ambiguity hypothesis posits that words perceived as highly emotionally ambiguous can retrieve more varied emotional contexts, which prompt enhanced elaborative processing that redounds to memory. Recent research has found supporting evidence that higher valence ambiguity leads to enhanced elaborative processing. However, it remains untested whether the enhanced elaborative processing is causally driven by higher variability in emotional contexts, as the emotional hypothesis assumes. To test this assumption, the current study experimentally manipulated the variability of emotional valence across encoding contexts. Specifically, participants studied real words (Experiment 1) or pseudowords (Experiment 2), each embedded in either two positive contexts (positive encoding), two negative contexts (negative encoding), or one positive and one negative context (mixed encoding). Their memory was then tested with free recall (Experiment 1) or speeded recognition (Experiment 2). Results showed that the mixed encoding condition yielded better recall for real words (Experiment 1) as well as better discrimination for pseudowords (Experiment 2) than the negative encoding condition. Additionally, in Experiment 1, the dual-retrieval model of recall revealed that the reconstruction parameter, which is sensitive to elaborative processing of semantic content, was significantly higher in the mixed encoding condition than both the positive and negative encoding conditions. These findings provide evidence that emotional context variability enhances elaborative reconstructive retrieval, aligning with the assumption of the emotional ambiguity hypothesis. More broadly, these findings offer new insights into how context shapes word retrieval: Lexical representations reflect not only distributional properties but also variability in the emotional experiences associated with words.