Most (but not all) quantifiers are interpreted immediately in visual context

Abstract

For our analyses, we excluded all participants in the final analyses with less than 50% data points in each of the experimental conditions. While this procedure affected eight participants when incorrectly answered trials were included, it affected an additional seven participants when incorrectly answered trials were excluded. Thus, the current additional ERP and reaction time analyses were carried out with n = 16 participants.

Context and Complexity in Incremental Sentence Interpretation: An ERP Study on Temporal Quantification

“…Importantly for the present considerations, in Augurzky et al (2020), positive and negative quantifiers elicited differential mid-sentence brain responses following the identical pictures, therefore ruling out a strategy according to which participants just focused on the pictorial information and mapped this information with the lexical contents of the color adjective (see also Augurzky et al, 2019, for a study on the scalar quantifier some, which indicates immediate quantifier effects on mid-sentence ERPs).…”

“…However, we believe that a unified account for the N400 effects across sentence positions, and also across experiments, may explain the whole range of data more parsimoniously than an account that assumes qualitatively different processes just for the infinitive position in the current experiment. Given the highly comparable ERP effects at early and later positions across studies on quantifier restriction, we consider it unlikely that ERPs at the infinitive and the PP noun in the present study need to be associated with different processes (see the comparable pattern at early and late sentence positions in Augurzky et al, 2016Augurzky et al, , 2019Augurzky et al, , 2020, which examine nominal quantifiers that differ about their semantic properties and pictures that differ concerning their complexity).…”

“…EEG data were analyzed statistically by carrying out repeated-measures ANOVAs for mean amplitude values within time windows per condition, which were chosen based on previous studies (Augurzky et al, 2017): a 300-400 time window was chosen to capture the N400, and a later 450-800 ms time window as used for capturing later effects. Based on previous studies on quantifier domain restriction (Augurzky et al, 2017(Augurzky et al, , 2019(Augurzky et al, , 2020, four regions of interest (Rois) were introduced into the analysis of variance: left anterior (Roi 1: F3, F7, FC1, and FC5), right anterior (Roi 2: F4, F8, FC2, and FC6), left posterior (Roi 3: CP1, CP5, C3, and P3), and right posterior (Roi 4: CP2, CP6, C4, and P4). From infinitive onset, the analysis involved the factors CONTEXT (A_complex locally false , B_simple true , C_simple false , D_complex locally false ), and ROI (1, 2, 3, 4), whereas from the preposition, the analysis included the factors CONTEXT (A_complex locally false , B_simple true , C_simple false , D_complex locally false ), TAGE (Arbeitstage, Feiertage), and ROI (1, 2, 3, 4).…”

“…Analogous to previous studies on quantifier restriction (Augurzky et al, 2017(Augurzky et al, , 2019(Augurzky et al, , 2020, we applied bandpass filtering without baseline correction. This method has become common practice in recent ERP studies on language comprehension (e.g., Bornkessel-Schlesewsky, Roehm, Mailhammer, & Schlesewsky, 2020;Chan et al, 2020).…”

“…A priming view, at least as we understand it, would highlight lexical processing strategies and, in an extreme version, it would even imply that the effects on the infinitive do not tell us anything at all about compositional-semantic processing. Augurzky et al (2020) summarize some of the most relevant arguments against such a purely lexical-driven strategy. Crucially, to arrive at a truth evaluation locally (in that study, positive and negative quantifiers were examined), readers need to syntactically combine the incoming sentence material and, at the same time, consider the semantic properties of the quantifier already for evaluating the information mid-sentence (in that study, the position of interest was a color adjective).…”