|

楼主 |
发表于 2005-4-3 17:47:25
|
显示全部楼层
Abstract
Language transfer has been a central issue in second language acquisition (SLA) for at least a century. As to when the transfer occurs, the early theories, such as Contrastive Analysis, Error Analysis and Interlanguage Hypothesis, fail to provide a satisfactory answer.
With the strong influences of Universal Grammar and Typological Universals, more researchers come to study transfer from the perspective of linguistic universals. One of the most famous and influential hypotheses is that “linguistically unmarked features of the first language (L1) will tend to transfer, but the linguistically marked L1 features will not” (Larsen-Freeman and Long 101). This hypothesis has been supported by many western studies, but in China, few researches have been conducted.
A contrastive analysis was carried out to see the similarities and differences between English and Chinese on the level of word order, clausal negation and relative clauses. Based on the typological criterion for determining the level of markedness, we assume that (1) the OSV order of Chinese declaratives is marked and less likely to transfer into interlanguage; (2) the preverbal negation of Chinese clause is unmarked and thus is submitted to transfer; and (3) the left branching relative clause in Chinese is a highly marked construction and thus is less likely to transfer. These serve as the working hypotheses of our study.
We thus design an acceptability judgment test and a translation test to testify these hypotheses. The acceptability judgment test consists of 30 English sentences, with word order, clausal negation and relative clauses being the three testing points. The respondents were required to make judgment on the acceptability of each experimental sentence as well as control sentence, and then grade them. The translation test consists of 15 Chinese sentences, which are expected to be translated into English. All the testing points are the same with those in the acceptability judgment test.
Our 120 subjects were from Henan University and its affiliated middle school. They were divided into 4 groups (30 in each): second year junior middle school students, first year senior middle school students, third year senior middle school students and college sophomores. Those 30 undergraduates and graduates from Clemson University in the U.S. formed the control group.
Data analysis involves two tasks: (1) As for grammaticality judgment test, we did the necessary arithmetical operations to describe and compare the grades for each testing point between different groups; (2) as for translation test, we conducted the error analysis and picked out the transfer errors.
The results of the two tests prove to be the same: (1) OSV Chinese declaratives will not transfer to English; (2) the preverbal feature of clausal negation in Chinese tends to transfer; and (3) the left-branching feature of Chinese relative clause is obviously subject to transfer.
In the above findings, (1) and (2) support the hypothesis, while (3) proves to be a counterexample. To the extent that marked structures are not expected to undergo transfer, the systematic transfer of pronominal relative clauses comes as a surprise. If our findings were proved valid by more researches, then the above hypothesis needs to be reconsidered—sometimes marked L1 features may also be transferred.
[ 本贴由 dogbiteyou 于 2005-4-3 17:51 最后编辑 ] |
|