Humboldt-Universität zu Berlin - Sprach- und literaturwissenschaftliche Fakultät - Institut für Slawistik und Hungarologie

10.1. Mariia Razguliaeva, Maria Onoeva, Radek Šimík & Roland Meyer (HU Berlin & Prague): Polar question meaning in Slavic: Visual world eye-tracking evidence [hybrid]

Polar questions (PQs) have traditionally been analyzed as denoting the set of their two possible answers (Hamblin 1973) or the corresponding bipartition of possible worlds (Groenendijk & Stokhof 1984). Setting important technical details aside, a PQ like "Is it raining?" would denote the set {it is raining, it is not raining}, or {p, ~p} with p=[[it is raining]]. This simplistic view has recently been challenged from various angles. First, there is growing evidence that the baseline semantics is crucially complemented by PQ bias - a preference for one of the two answers, based either in the speaker's beliefs or in the contextual evidence (Büring & Gunlogson 2000; Romero & Han 2004; Sudo 2013; Goodhue 2022; a.o.). Second, some authors have questioned the very idea that PQs denote a set of two propositions. Biezma & Rawlins (2012) argue that a polar question denotes a singleton set {p} with the relevance of the other alternative being merely implicated.

We aim to shed light on the meaning of PQs by investigating their online processing using an eye-tracking visual world experiment. Building on Tian et al.'s (2016, 2021) work on English and French, we look at the processing of Russian and Czech PQs and assertions. We ran a number of closely parallel visual world eye-tracking experiments in the two language mutations. All stimuli were short prerecorded dialogues starting with the target utterance containing a transitive verb, its subject and object, and an adverbial, followed by a brief response of the form `I think that yes/no'. In the target utternace, we manipulated several variables: 1) force (question vs. statement); 2) polarity (affirmative vs. negative); 3) verb position (sentece-initial or medial); 4) presence or absence of a particle (razve in Russian, snad and asi in Czech). Simultaneously with the auditory linguistic stimulus, participants were presented with four pictures on a screen, two corresponding to the two alternatives p and ~p, and two showing unrelated distractors. We measured the durations of fixations to p and ~p images during different parts of the utterance. Our results corroborate Tian et al.'s findings with new evidence from Slavic languages, and provide support for Biezma & Rawlins' (2012) singleton-based analysis of PQs, as well as preliminary evidence for the presence of positive epistemic bias in negative PQs.