Neural correlates of visualizations of concrete and abstract words in preschool children: a developmental embodied approach

Abstract

The neural correlates of visualization underlying word comprehension were examined in preschool children. On each trial, a concrete or abstract word was delivered binaurally (part 1: post-auditory visualization), followed by a four-picture array (a target plus three distractors; part 2: matching visualization). Children were to select the picture matching the word they heard in part 1. Event-related potentials (ERPs) locked to each stimulus presentation and task interval were averaged over sets of trials of increasing word abstractness. ERP time-course during both parts of the task showed that early activity (i.e., <300 ms) was predominant in response to concrete words, while activity in response to abstract words became evident only at intermediate (i.e., 300-699 ms) and late (i.e., 700-1000 ms) ERP intervals. Specifically, ERP topography showed that while early activity during post-auditory visualization was linked to left temporo-parietal areas for concrete words, early activity during matching visualization occurred mostly in occipito-parietal areas for concrete words, but more anteriorly in centro-parietal areas for abstract words. In intermediate ERPs, post-auditory visualization coincided with parieto-occipital and parieto-frontal activity in response to both concrete and abstract words, while in matching visualization a parieto-central activity was common to both types of words. In the late ERPs for both types of words, the post-auditory visualization involved right-hemispheric activity following a "post-anterior" pathway sequence: occipital, parietal, and temporal areas; conversely, matching visualization involved left-hemispheric activity following an "ant-posterior" pathway sequence: frontal, temporal, parietal, and occipital areas. These results suggest that, similarly, for concrete and abstract words, meaning in young children depends on variably complex visualization processes integrating visuo-auditory experiences and supramodal embodying representations.

Keywords: ERPs; embodied cognition; preschool children; visual mental imagery; visualization; word processing.

FIGURE 1

Representation of the word verification task (WVT) used in the present experiment, indicating stimulus timing for the auditory and visual parts of the task. The interval for the auditory and visual averages was -200 to 1000 ms around stimulus onset followed by a correct response to the visual stimulus. The visual presentation remained on the screen until the participant made a response. The length of the post-response cross fix allowed the researcher to re-iterate the experiment requirements to the participant (e.g., “try not to blink, remain still,” etc.) if necessary.

FIGURE 2

(A) Study of waveform differences and homologies empirically observed for ERP activity in response to concrete and abstract words. To quantify distances shown in the waves, the Iso-polarity effects involved computing the differences in absolute values, whereas reverse polarity effects involved computing the sum of absolute values. The significance threshold shown as red distance-bar was corrected for multiple comparison testing using the Simes–Bonferroni procedure (see text). Grand average ERP waveforms for concrete and abstract words conditions in the auditory (B) and visual parts (C) of the WVT. The magnitude of the minimum significant effect is represented as a red distance-bar (drawn in scale with respect to 5 μV). Distances/differences larger than the red bar were significant at least at p < 0.05, as examples, main points of significance are marked with the red bar in the waveforms.

FIGURE 3

Event-related potential activity pathways analysis summarizing the spread (least-path connections) of observed scalp grand average ERP activity from one electrode/area to the next. All correlations reported (shown in blue) were significant (p < 0.05) after multiple comparison correction.