Search Abstracts | Symposia | Slide Sessions | Poster Sessions
How emotion influences prediction dynamics during sentence reading: Evidence from brain potentials
Poster Session A - Sandbox Series, Thursday, October 24, 10:00 - 11:30 am, Great Hall 3 and 4
This poster is part of the Sandbox Series.
Jing-Yi Huang1, Yi-Chun Ko1, Joshua Oon Soo Goh1, Chia-Lin Lee1; 1National Taiwan University
Emotions shape human experiences and interpersonal communication. In language comprehension, emotional context can heighten attention allocation and elaborate re-evaluation of incoming stimuli. Prior research indicated that more positive amplitudes were elicited in unexpected plausible neutral continuations following emotional contexts than emotionally predicted endings in young adults. This effect is frontally distributed and post-N400, thus is reminiscent of the anterior positivity (AP) effect in response to violations of strong predictions in language comprehension studies that did not systematically manipulate emotion. These results thus indicate a possibility that emotional contexts may enhance predictive processing during sentence reading. The present study aims to further elucidate the nature of this anterior effect following emotional contexts. As these prior findings only used emotionally unexpected endings in moderately strong constraint conditions, the anterior response differences can alternatively be due to anterior negativity (AN) to emotionally expected predictions compared to emotionally unexpected endings, which is not yoked to predictions. To elucidate these possibilities of the anterior effect in an emotional context, we manipulated Sentence Constraint (strong: 85% vs. weak: 20%), Ending Word Expectedness (expected: 20-85% vs. unexpected: 0%), and Context Valence (emotional: 2.2 vs. neutral: 0.4, mean differences from the midpoint 5 on a 1-9 scale). Emotional contexts included equal numbers of positive and negative scenarios. All unexpected sentence-final words were neutral and all sentences were plausible. Target words were controlled for familiarity, concreteness, and length. 480 sentences were arranged into 2 lists, with each condition consisting of 30 sentences. Each participant is assigned to 1 list and reads each sentence context only once. Participants will include 30 young native Taiwan Mandarin speakers. Each read sentences word-by-word on a computer monitor while Event-Related Potentials (ERPs) were recorded. Sentence recognition questions will be posed intermittently to ensure engagement. We hypothesize validation of expectancy and constraint manipulations through graded N400 effects, with unexpected words producing the most negative N400s and expected words showing a graded reduction, more so for strongly constraining sentences than weakly constraining ones. We expect to replicate AP effects with neutral contexts: unexpected words in the strong constraint condition should elicit more pronounced AP compared to expected words in strong constraint condition, along with both expected and unexpected words in weak constraint conditions. Our critical analysis focuses on emotional sentences to determine if unexpected endings in strong constraint conditions elicit the AP effect or if strong expected endings in strong constraint elicit the AN compared to the other three emotion conditions. With this design, this study aims to elucidate these possibilities of the anterior effect in an emotional context and the interaction between emotion and language processing, highlighting the potential role of emotion in enhancing predictive language comprehension.
Topic Areas: Reading, Control, Selection, and Executive Processes