Poster Session C (Saturday 12:10-2:00 PM) I am happy to answer questions via email at mbh32@pitt.edu or OSF comments, or live during the poster session via a Zoom meeting from 12:00-12:45 PM ET (https://zoom.us/j/6348425095 Meeting ID: 634-842-5095) Abstract: This study provides novel neuroimaging evidence that comprehenders make both coarse-grained featural and fine-grained lexical predictions, and that these predictions serve to facilitate processing. Participants read plausible SVO sentences up to the object’s determiner and judged whether the upcoming noun, presented following a variable ISI, would be animate or inanimate. Multivariate analyses revealed that animacy-sensitive regions (and left SMG) successfully classified sentences that predicted specific animate themes. Univariate analyses showed reduced BOLD response in animacy-sensitive regions for general-animate-theme-predicting sentences. This pattern appeared both before and during the noun, suggesting that predictive sentence contexts facilitated comprehenders’ processing of upcoming (animate) themes.