Poster B56, Thursday, August 16, 3:05 – 4:50 pm, Room 2000AB

ERP correlates of picking up new foreign-language words in dialogue

Kristin Lemhöfer1, Julia Egger1,2, James McQueen1;1Radboud University, 2Max-Planck-Institute for Psycholinguistics

A lot of applied research has been devoted to identifying suitable methods to teach people new words in a second language (L2). However, many (especially immersed) L2 speakers improve their vocabulary spontaneously and ‘in the wild’, i.e. in natural communicative situations, rather than in the classroom. Investigating this kind of word learning, also termed ‘incidental learning’, in the laboratory has proven difficult, especially with respect to concealing the learning character of the study from participants. In a recent line of research, we developed an experimental method of incidental L2 word learning in a dialogue-like situation that turned out to be highly successful in that respect. This paradigm entails that the participants are made to believe that the study is about price judgments and their consistency. In the current study, we adapted this paradigm for an ERP context, to investigate the electrophysiological signatures of picking up new words from a dialogue partner. Native Dutch speakers first completed a pre-test on the experimental set of target and filler objects in their L2 English, supposedly because the experimenter did not speak Dutch. They had to name the objects and to give a price estimate (e.g., “a fridge costs 200 Euros”). Unbeknownst to the participants, the real aim of this pre-test was to test initial word knowledge on the experimental target items which were typically unknown to this L2 population (e.g., whisk), while the fillers were typically known (e.g., fridge). After that, participants alternately made price comparisons of pairs of depicted objects (e.g., “a whisk is cheaper than a fridge”), or listened to and judged those by a virtual partner (voice recordings). Critically, the target items were always produced by the virtual partner first, providing learning-relevant input to the participant. After a fixed number of trials, the same object re-appeared in the participant’s trial, revealing whether or not she had learned the word from the input. In a second block, all words were presented in the same fashion again to provide a second opportunity for learning. As in our previous studies, only few participants suspected the study to be about word learning. On average, 36% of the previously unknown words were learned after the first, and another 20% after the second exposure. In the ERPs, we failed to find the expected larger N400 for unknown (target) vs. known (filler) words that is normally observed for pseudowords in sentences, and found a late positivity instead. This suggests that new to-be learned words in a meaningful context are encoded differently than meaningless pseudowords. Furthermore, there was a late positivity for later learned vs. forgotten words, showing that the neural processes during encountering a new L2 word in a natural situation are predictive of subsequent learning. Finally, newly learnt words became indistinguishable in their ERP signature from previously known filler words in the second block, i.e. after only one exposure, indicating that incidental word learning can operate surprisingly fast. This study is the first to reveal neural correlates of incidental learning and memory for new L2 words.

Topic Area: Multilingualism

Back