Emotional representations of a word depend on different sentential contexts. For instance, ?holiday? feels pleasant when it appears as a single word. But when it is embedded in a sentential context, e.g., ?holidays are stressful?, the emotional connotation of the word can be instantaneously changed into something negative. This project investigates the effect of context on the emotional connotations of words in older adults, in two ways. First, older adults typically show a positivity bias--i.e., the tendency to focus on positive information. How do older adults update the emotional connotations of words in positive or negative contexts, compared with younger adults? Second, in emotionally ambiguous scenarios, would older adults predict positive endings, as a result of their positivity bias? These findings shed light on the link between sentence processing and emotion regulation, and how such age-related positivity bias can be leveraged to improve older adults? emotional well-being, which is a pressing issue in this aging society. This project uses web-based surveys and electroencephalography (EEG) to investigate how the emotional representation of a word is retrieved, updated, and pre-activated during language comprehension in older and younger adults. The overarching hypothesis is that age-related positivity bias influences the affective features of a word, not only during the reading of the target word given its preceding context, but also the prediction or predicted features prior to reading the target word. Two studies are conducted. One study examines whether and how age-related positivity bias influences the updating of emotional features of a word during sentence comprehension. Target words are embedded in positive or negative sentences, e.g., "Holiday is stressful/wonderful", first and then are presented again, e.g., "holiday". Brainwave analyses are performed at the first and the second occurrences of the target word, to identify changes of emotional representations of the word. The expected outcome is that older adults will be more effective in contextual updating in a positive context than in a negative context, in comparison to younger adults. The other study investigates whether and how age-related positivity bias influences prediction of affective features in emotionally ambiguous sentences, e.g., "Mary received the exam results". Participants read such sentences first, and then an emotionally congruent or incongruent target word, e.g., "pass" or "fail". The expected outcome is that older adults will pre-activate positive features congruent with the positive target, but incongruent with the negative target. These results will contribute to emotional language comprehension and predictive mechanisms in language across the lifespan. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.