Research shows that language processing mechanisms are permeable to the speaker’s accent, but virtually no data exists on non-literal language. Our online rating study investigated whether accent-based biases could hinder making inferences from ironic speech. Ninety-six participants listened to dialogues between native and foreign-accented English speakers and rated them on several scales. We found that the ironic intent in the accented speech was missed significantly more often than in the native speech for all irony types. Importantly, participants’ individual differences significantly affected the ratings and interacted with both accent and irony-type. More conservative participants were worse at detecting irony than their liberal peers but this effect was stronger for accented speech and a rarer irony type. In contrast, high empathy facilitated irony detection. The results demonstrate that interpersonal variation in personality and moral values affects language comprehension and needs to be accounted for in models of language processing.

When people read the sentence “The babysitter put on a TV show for the kids because he needed to use the washroom,” the male identity prompted by the pronoun clashes with the stereotypical expectation of babysitters as female, rendering the pronoun “he” more difficult to process than “she” would be. We ask whether participants’ HEXACO PI-R personality traits, Political Ideology, and Disgust Sensitivity (DS-R) modulate their reactions to pronouns Congruent versus Incongruent with stereotyped role nouns. 80 English-speaking participants read 40 sentences with female/male stereotypes and were asked to rate each item a Likert scale (1-6) from “Completely Inappropriate” to “Completely Appropriate.” Initial analysis indicates that Openness correlates with higher ratings of Appropriateness for Incongruent stereotypes, and Introversion correlates with low Appropriateness ratings of Incongruent, Female stereotypes. We also expect a correlation between high Conservativeness, Disgust Sensitivity and low ratings of Appropriateness for Incongruent stereotypes.

Recent research suggests that an individual's disgust sensitivity affects language comprehension and correlates with political attitudes. Importantly, disgust sensitivity is not a stable measure and can be manipulated dynamically. We investigated the effect of the ongoing pandemic on language processing in a word rating and lexical decision study. Each participant was first exposed to either headlines portraying CoViD-19 as a serious disease or those downplaying it. The results showed an interaction between person-based factors, inherent word characteristics, and the participants’ responses. After reading headlines emphasizing the threat of CoViD-19, easily disgusted participants considered the least disgusting words more disgusting. Further, political views played a role. More liberal participants rated the words lower for disgust in the downplayed condition but higher in the severe condition than their more conservative peers. The results of this study shed new light on how the media's stance on the pandemic may affect the public’s response.

Using a novel adaptation of the visual world eye-trackingparadigm we investigated children’s and adults’ onlineprocessing of reference in a naturalistic language context.Participants listened to a 5-minute long storybook whilewearing eye-tracking glasses. The gaze data were analyzedrelative to the onset of referring expressions (i.e., full nounphrases (NPs) and pronouns) that were mentionedthroughout the story. We found that following the mentionof a referring expression there was an increase in theproportion of looks to the intended referent for both childrenand adults. However, this effect was only found early on inthe story. As the story progressed, the likelihood thatparticipants directed their eye gaze towards the intendedreferent decreased. We also found differences in the eye gazepatterns between NPs and pronouns, as well as betweenchildren and adults. Overall these findings demonstrate thatthe mapping between linguistic input and corresponding eyemovements is heavily influenced by discourse context.

We use visual world eye-tracking to test if a speaker’s eyegaze to a potential antecedent modulates the listener’sinterpretation of an ambiguous pronoun. Participants listenedto stories that included an ambiguous pronoun, such as “Thedolphin kisses the goldfish... He....” During the pre-pronominal context, an onscreen narrator gazed at one of thetwo characters. As expected, participants looked more at thesubject character overall. However, this was modulated by thenarrator’s eye gaze and the amount of time the participantspent looking at the gaze cue. For trials in which participantsattended to the narrator’s eye gaze for > 500ms, participantswere significantly more likely to interpret the pronoun asreferring to the object if the narrator had previously looked atthe object. Results suggest that eye gaze – a social cue – cantemper even strong linguistic/cognitive biases in pronounresolution, such as the subject/first-mention bias.

The present study investigates individual differences in pupildilation during standard word naming. We looked at (i) howindividual subjects’ pupil size changes over the course of timeand (ii) how well pupil size is predicted by the frequency ofthe stimuli. The time course of the pupil size was analysedwith generalized additive modeling. The results show large in-dividual variations in the pupil response pattern in this verysimple task. Although, we see a pupil response to both stimu-lus onset and articulation onset and offset, both the amplitudeof change and the direction of change differ substantially be-tween subjects. This raises the question of what makes thepupil response functions so diverse, and one factor indicatedby the frequency effect or the lack thereof might be shallowreading versus reading for content.