Hi all! Thanks for stopping by! I will be active via OSF ( during **Poster Session A**) to answer any questions about my presentation. If you run out of time and/or for some reason cannot reach me online, all questions can also be directed to my email address: elyceely@live.unc.edu How does language experience affect processing mechanisms? It is well established that frequency of exposure affects lexical and syntactic processing (e.g., MacDonald et al., 1994; Trueswell & Tanenhaus, 1994). But little is known about exposure effects on discourse processing. We examine this question by testing individual differences in the use of implicit causality (IC) to interpret pronouns. E.g., in Matt feared Will because he..., or Will scared Matt because he…, people tend to assume that Will is the likely cause of both events, and prefer to assign “he” to Will (Garvey and Caramazza, 1974; Hartshorne et al., 2015; Kehler & Rohde, 2013). This bias occurs alongside a general preference to assign pronouns to the subject character (Stevenson et al., 1994). Thus, in these structures people follow both syntactic and semantic cues to interpret pronouns. Do individuals differ in how much they weight each cue? Recent evidence shows that print exposure (aka reading) can affect pronoun interpretation biases. In sentences like Ana is cleaning up with Liz. She needs the broom, people prefer the subject character (Ana) as the pronoun referent. However, this bias is stronger for individuals with greater print exposure (Arnold, Strangmann, Hwang, Zerkle, & Nappa, 2018). Print exposure affected interpretation in spoken language, so this effect is unrelated to reading skill. A similar effect occurs for transfer events like Ana threw the ball to Liz and then she... or Liz received the ball from Ana and then she…. people tend show a bias to assign the pronoun to the goal character (Liz), as well as a general subject bias. Yet print exposure only correlates with the strength of the subject bias, not the goal bias (Langlois & Arnold, in press). This raises questions about how people use IC during pronoun comprehension. Does print exposure modulate this effect at all, or does print exposure only modulate the preference for assigning pronouns to the subject character? Methods. We test this question by measuring individual patterns of pronoun comprehension in IC contexts, and comparing these with print exposure scores. We measured print exposure with a proxy measure, the Author Recognition Task (ART, Stanovitch & West, 1989). Participants identify real authors out of a list of real and fake authors; and scores are known to correlate with other measures of language knowledge (Mani & Huettig, 2014; Montag & Macdonald, 2015). We measured spoken pronoun comprehension in two tasks. In the Nonword Task (Exp. 1; Hartshorne & Snedeker, 2013), participants heard implicit cause (IC) stories that ended in a novel word i.e. dax. In the Congruency Task (Exp. 2), participants heard stories where the pronoun clause was more congruent with the IC interpretation, (see Table 1). The pronoun always matched the gender of both characters. Participants were asked two comprehension questions, one of which probed the pronoun interpretation (see Table 1). The Nonword task assesses biases that stem from the context, while the Congruency task tests sensitivity to the semantic congruency of the final clause with the IC context. Verb type was manipulated, so the IC was the subject in half the stories and the object in the other half. Critical stimuli were intermixed with non-IC filler sentences that had two different-gender characters. Our dependent measure was selection of the grammatical subject character. Each experiment had 60 participants. Results: Participants tended to identify the IC as the referent, selecting the subject more when it was the IC than not (Exp. 1 & Exp. 2, p <.0001). Critically, ART interacted with Verb bias (Exp. 1 p =.0002; Exp. 2, p =.0014), such that as ART score increased, participants were more likely to follow a semantic bias (see Figs. 1 and 2). Conclusions: Print exposure increases sensitivity to implicit causality during pronoun comprehension. In Exp. 2 this could reflect a better ability to choose the contextually-appropriate response. But in Exp. 1, both interpretations are “correct”. This suggests that exposure affects predictions based on the verb and coherence relation. Current models suggest that implicit causality guides representations of reference probability (Hartshorne et al., 2015; Kehler & Rohde, 2013). Our finding extends these models, suggesting that the calculation of these probabilities is supported by language exposure.