Case Study: How Linguistic Experience Influences Predictive Sentence Processing

Background image

Researchers of the Max Planck Institute for Psycholinguistics in Nijmegen used SMI Remote Eye Tracking to study the processing of semantic cues for prediction among native and second language (L2) users of Dutch using a visual world paradigm.

 

Challenge

Investigating the role of experience for prediction in written and spoken sentence processing in adults has in the past involved adults with dyslexia or low literacy, compared to non-dyslexics and literates.

The challenge of this study was to compare how experience with language-specific verbal semantics modulates predictive sentence processing of advanced second language users (L2) with different native language (L1) backgrounds.

Solution

Researchers investigated the differences between native Dutch speakers (L1) and German, French and English learners of Dutch (L2) when processing sentences describing placement events. Eye tracking was used to examine anticipatory eye movements resulting from linguistic cues in the placement verb.

This is interesting due to the differences between French / English and German / Dutch in describing those actions (put / mettre vs. zetten-leggen-plaatsen / hinstellen-hinlegen). During the experiments, participants were asked to listen to short sentences describing placement events while looking at objects placed in different positions (standing & lying).

SMI Eye Tracking

Eye movements of participants were recorded with SMI’s Remote Eye Tracking system at 500Hz sampling rate in order to capture fixations in the extremely short time windows of stimulus onsets (predictor verb zetten/legen and object revealing noun). For the analysis of the results,

SMI’s BeGaze software was used to define areas of interest (AOIs) to investigate the difference in the proportion of fixations on the displayed objects by native Dutch listeners and German, French and English natives.

Benefit

The analysis of the verbal prediction time window (200ms after verb onset until 200 ms after start of the noun) clearly showed that French/English Dutch L2 users lack the prediction effect of placement verb semantics.

Based on the cross-linguistic comparison of L2 processing, the researchers concluded that prior experience with a specific semantic contrast in one’s L1 facilitates prediction in L2, and hence adds to the evidence that linguistic experience modulates predictive sentence processing.

Background

Language comprehenders can use linguistic and nonlinguistic cues to actively anticipate upcoming words and structures within a sentence. The capacity to generate predictions is considered crucial to language processing.

A different way of investigating the role of linguistic experience in predictive language processing is by studying second language (L2) learners’ predictive ability in their L2. The rationale here is that native and non-native/ L2 speakers differ in their experience with the concepts and forms encoded in the second language.

Benefit

The analysis of the verbal prediction time window (200ms after verb onset until 200 ms after start of the noun) clearly showed that French/English Dutch L2 users lack the prediction effect of placement verb semantics.

Based on the cross-linguistic comparison of L2 processing, the researchers concluded that prior experience with a specific semantic contrast in one’s L1 facilitates prediction in L2, and hence adds to the evidence that linguistic experience modulates predictive sentence processing.

Customer institution Max Planck Institute for Psycholinguistics, Nijmegen, Netherlands
Customer website Link
Related publication