Cross-script cognate priming effects on visual word recognition:Effects of Japanese loanword cognates in L2 Japanese learners

Abstract

Research in bilingualism has shown that translation pairs that look or sound similar across languages (e.g., English-Spanish rich-rico) are easy to recognize for the speakers of the two languages. Such translation pairs are called cognates and the processing advantage of cognates is known as cognate facilitation effect. This thesis investigated cognate facilitation effect in visual word recognition by examining masked priming effects in beginning English-Japanese bilinguals, namely late second language (L2) learners of Japanese whose first language (L1) is English. More specifically, the current study examined (i) what constitutes cognate facilitation effect in visual word recognition in bilinguals whose two languages share no orthographic similarity and (ii) whether the facilitation effect would be modulated by L2 proficiency by testing two proficiency groups of L2 learners of Japanese. The results from masked priming experiments using English (L1) primes with Japanese (L2) targets showed that priming effects for cognate pairs (e.g., card-カード/ca:do/) were larger than for translation pairs (e.g., desk-つくえ/tukue/) and the effect was smallest for word pairs similar only in sound (e.g., nail-メール/me:ru/,mail). The same patterns of results were obtained for Japanese (L2) primes with English (L1) targets. The size of priming effects did not differ across the two proficiency groups, but significantly larger priming effects were observed in L1-L2 priming direction than in L2-L1 priming direction. These findings indicate that (i) cognate facilitation effect in visual word recognition can be obtained without shared orthography, and shared semantics and phonology underlie the cognate facilitation effect. Further, (ii) the asymmetry in the size of the priming effects in beginning bilinguals may be due to their low L2 proficiency and the fact that sufficient L2 proficiency is required to utilize cognate information from L2 primes in the process of recognizing L1 targets. The findings are discussed in regard to cross-language co-activation and interaction during bilingual lexical processing within the framework of the Bilingual Interactive Activation (BIA+) model. The role of L2 proficiency in bilingual lexical processing is also considered.