Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Lightning Talks
The impact of mirative markers on reaction to unexpected words during sentence reading
There is a Poster PDF for this presentation, but you must be a current member or registered to attend SNL 2023 to view it. Please go to your Account Home page to register.
Poster C117 in Poster Session C, Wednesday, October 25, 10:15 am - 12:00 pm CEST, Espace Vieux-Port
Benjamin Menashe1, Hadar Altshuler-Frenkel2, Yael Greenberg2, Michal Ben-Shachar1,2; 1The Gonda Multidisciplinary Brain Research Center, Bar-Ilan University, Israel, 2Department of English Literature and Linguistics, Bar-Ilan University, Israel
During sentence processing, unexpected words typically evoke neural and behavioral responses, such as longer reading times, compared to expected words. Mirative markers, e.g., “surprisingly”, are elements which linguistically encode a violation of expectation, and can be used for expectation-management during discourse (Delancey, 1997). The present study examined the modulatory effect of mirative markers on the response to semantically unexpected words presented in sentential context. 60 native-Hebrew speakers (48F:12M, 18 to 34 years old) with no diagnosed ADHD or dyslexia read 160 Hebrew sentences, in a self-paced reading paradigm. Participants controlled word-by-word presentation with button presses, and their reaction time to each word was recorded. We manipulated, in a 2X2 design, the presence of a mirative marker in the beginning of the sentence, and the semantic expectedness of a target word that appeared downstream. Specifically, sentences in the Mirative condition began with “I was surprised that”, while sentences in the Neutral condition began with “I saw that” (in Hebrew these terms both consist of a single word, 3 syllables long). Sentences were all of SVO form. The target word appeared in the 6th or 7th position in the sentence, and was followed by 3 more words until sentence termination, to avoid wrap up effects. Target words were matched for mean length between conditions. Sentences were chosen using a pre-test questionnaire with 20 respondents (none of whom participated in the study), who rated the expectedness of the target word in a 1-7 Likert scale. In the self-paced reading experiment, 20% of the sentences were followed by a 2-alternative forced-choice question in order to ensure participant comprehension. A linear mixed-effects model was fit to predict the log-transformed reaction time at target words, with fixed effects of expectedness and mirativity, random intercepts and slopes per participant and per sentence, and target word frequency and sentence number as control covariates. We found a significant interaction between expectedness and mirativity (beta=-0.05, se=0.02, p<0.01), such that the presence of a mirative marker decreased the RT for unexpected target words and increased RT for expected target words. Within the Neutral condition, but not within the Mirative condition, we found a significant simple effect of unexpectedness (beta=0.04, se=0.01, p<0.001). Additionally, we found a significant main effect of unexpectedness at the word following the target (“spillover effect”) (beta=0.03, se=0.01, p < 0.01). We conclude that mirative markers prepare readers for an upcoming surprise, thereby modifying their expectations and leading to a reduction in their surprise-like responses to later events. This may be achieved by reducing the prediction error which an unexpected word usually elicits, thus lowering the processing cost. However, such a benefit may be limited in time, as evidenced by the longer RT for unexpected words in the "spillover" zone, even when the sentence was introduced by a mirative marker. Future planned studies will further examine the dynamics of expectation management and investigate the neural effects of mirative markers.
Topic Areas: Reading, Meaning: Discourse and Pragmatics