Abstract
Previous studies have shown that in second language vocabulary development, language
learners tend to connect an L2 form to an L1 equivalent translation by transferring
entire L1 semantic and syntactic knowledge to L2. This tendency can be observed even
when two L2 lexical items share only one L1 translation. As a result, second language
learners may reveal usages diverse from those of native speakers and/or be less capable
of differentiating the nuance of meanings of a word pair which share one and the same
L1 translation equivalent. For example, given that question and problem are both
translated into wèntí (問題)in Chinese, Chinese learners of English are liable to
make a linguistic mistake such as *I have a problem to indicate I have a question.
Previous research has used cloze tests and measurements of reaction time to
demonstrate the psycholinguistic phenomenon of lexical transfer. However, whether
reliance on L1-mediated concepts or on mental links established between L1 words and
L1-mediated concepts contributes to the transfer is not clear, as the former results in
conceptual transfer and the latter semantic transfer.
The present study investigates this crosslinguistic influence (lexical transfer) with a
different methodology. Four word pairs each sharing one and the same Chinese
translation were selected from previous studies: criterion-standard (both are translated
to ‘biāozhǔn’ 標準 in Chinese), complicated-complex (‘fùzá’ 複雜), safe-secure
(‘ānquán’ 安全) and doubt-suspect (‘huáiyí’ 懷疑). Six Chinese learners at the same
level of English proficiency were involved in this longitudinal study. They participated
in a series of group guided conversations; a guided conversation was held twice for each
word pair and video recorded. Following the guided conversation, a retrospection
procedure took place individually with each participant in the learner’s L1 to clarify the
learner’s use and understanding of the target words in previous conversation sessions.
The primary focus of the study is to investigate whether the target words were
interchangeably used in a fully contextual environment.
Results show that these learners may or may not transfer the entire L1 semantic and
syntactic knowledge to the L2 in real-time conversation. The type of lexical transfer
depends on the learner’s existing concepts and word meanings. Semantic transfer did
commonly happen when two L2 lexical items shared one L1 translation, but it did not
guarantee the learner’s interchangeable use of the target words. The retrospection
demonstrates that separated concepts could be expressed with a different sentence
pattern (without using a target word), or with a different syntactic structure (with a
target word or other words) in the learner’s L1. Thus two separate concepts associated
to a word pair may have already existed in the learner’s mind and consequently enabled
them to perceive the differences within a word pair. On the other hand, an aggregation
of non-distinguished concepts for a word pair may also exist in the learner’s mind,
which impedes differentiation of a pair. Hence semantic transfer and conceptual transfer
in second language vocabulary learning need to be carefully considered in that the
former is conceivable and readily avoided but the latter is hardly recognized and
reconstructed.