Orthographic Activation in L2 Spoken Word Recognition Depends on Proficiency: Evidence from Eye-Tracking

Analysing data from the psycholinguistic visual-world paradigm: Comparison of different analysis methods

In this paper, we discuss key characteristics and typical experimental designs of the visual-world paradigm and compare different methods of analysing eye-movement data. We discuss the nature of the eye-movement data from a visual-world study and provide data analysis tutorials on ANOVA, t-tests, linear mixed-effects model, growth curve analysis, cluster-based permutation analysis, bootstrapped differences of timeseries, generalised additive modelling, and divergence point analysis to enable psycholinguists to apply each analytical method to their own data. We discuss advantages and disadvantages of each method and offer recommendations about how to select an appropriate method depending on the research question and the experimental design.

The Role of Orthography in Lexical Processing of the Phonological Variants in Second Language

There is evidence that orthographic knowledge can influence on-line spoken-word recognition. Interestingly, when graphic and phonetic codes are not congruent due to the application of phonological alternation processes, people report hearing sounds that are matched to graphic (underlying), not phonetic codes (Hallé et al. in J Mem Lang 43:618-639, 2000). It is, however, not known whether the same effect arises in the processing of a non-native language (L2). In the present study, advanced Mandarin learners of Korean as well as native Korean listeners performed a phoneme monitoring task using words undergoing obstruent nasalization in Korean. The results showed that orthographic information dominated the phonetic judgments of the native Korean listeners, while the Mandarin learners' judgments relied more on the phonetic input. These results suggest that even the lexical access of highly experienced L2 learners differs from that of native speakers and that advanced learners still have difficulty employing orthographic information to access the L2 lexicon.

Everyday Language Exposure Shapes Prediction of Specific Words in Listening Comprehension: A Visual World Eye-Tracking Study

We investigated the effects of everyday language exposure on the prediction of orthographic and phonological forms of a highly predictable word during listening comprehension. Native Japanese speakers in Tokyo (Experiment 1) and Berlin (Experiment 2) listened to sentences that contained a predictable word and viewed four objects. The critical object represented the target word (e.g., /sakana/; fish), an orthographic competitor (e.g., /tuno/; horn), a phonological competitor (e.g., /sakura/; cherry blossom), or an unrelated word (e.g., /hon/; book). The three other objects were distractors. The Tokyo group fixated the target and the orthographic competitor over the unrelated objects before the target word was mentioned, suggesting that they pre-activated the orthographic form of the target word. The Berlin group showed a weaker bias toward the target than the Tokyo group, and they showed a tendency to fixate the orthographic competitor only when the orthographic similarity was very high. Thus, prediction effects were weaker in the Berlin group than in the Tokyo group. We found no evidence for the prediction of phonological information. The obtained group differences support probabilistic models of prediction, which regard the built-up language experience as a basis of prediction.

Native speakers like affixes, L2 speakers like letters? An overt visual priming study investigating the role of orthography in L2 morphological processing

In an overt visual priming experiment, we investigate the role of orthography in native (L1) and non-native (L2) processing of German morphologically complex words. We compare priming effects for inflected and derived morphologically related prime-target pairs versus otherwise matched, purely orthographically related pairs. The results show morphological priming effects in both the L1 and L2 group, with no significant difference between inflection and derivation. However, L2 speakers, but not L1 speakers, also showed significant priming for orthographically related pairs. Our results support the claim that L2 speakers focus more on surface-level information such as orthography during visual word recognition. This can cause orthographic priming effects in morphologically related prime-target pairs, which may conceal L1-L2 differences in morphological processing.

The Pre-attentive L2 Orthographic Perception Mechanism Utilized by Bilinguals with Different Proficiency Levels

Language proficiency is predicted to modulate orthographic-semantic association in second language (L2) vocabulary acquisition, in accordance with the assumptions of the Developmental Bilingual Interactive-Activation model (BIA-d) (Grainger et al., 2010). The current study explored this modulation during pre-attentive L2 orthographic perception. ERPs were recorded from Chinese–English bilinguals with different L2 proficiency during their pre-attentive response to deviant and standard stimuli arranged in the oddball paradigm. Two stimulus types were investigated separately: L2 orthography and L1 orthography. In the L2 orthography condition, a MMN-N400 complex (i.e., deviancy effect) was found in the high L2 proficiency bilinguals, but only a marginally significant reduced negativity in an early time window was found in the low L2 proficiency bilinguals. In the L1 orthography condition, the high and low L2 proficiency bilinguals showed similar deviancy effect in the form of MMN-P3a-LPC complex. The current findings suggest that proficiency modulates pre-attentive L2 orthographic perception, such that the high L2 proficiency bilinguals activate the associated semantic representation instantly upon orthographic decoding, while the orthographic-semantic connection is not activated for the low L2 proficiency bilinguals. This is probably due to their difference in the strength of orthographic-semantic association. These findings contribute to the understanding of orthographic processing by bilinguals at the pre-attentive level and provide supporting evidence for the BIA-d model.