The issue of pro-drop phenomenon concerning the availability of UG in SLA has been broadly discussed and debated in many languages, such as [+null subject] and [-null subject] languages, and issues concerning whether L2 learners are able to reset a parameter from the L1 value to the L2 value. If parameter resetting does occur, this can be viewed as the evidence for the UG access. In general, there are three kinds of languages which present different pro-drop parameters. Italian-like languages, such as Italian and Spanish, are considered as the [+null subject] languages. The rich subject-verb agreement systems enable them to drop pronouns, and pro can appear in the specifier position of any finite IP. Chinese-like languages, such as Chinese and Korean, also serve as the [+null subject] languages; however, the distinction between Chinese-like languages and Italian-like languages is the rule of Topic NP Deletion (Tsao, 1977) in Chinese-like languages. The rule operates across discourse to delete the topic of a sentence under the identification with a topic in a topic chain. Example (1) shows a topic chain that explains Tsao’s Topic NP Deletion (1977: 92):
(1) Nei ke shui, ei hua xiao, ei yezi da, eihen nakan, (suoyi) wo mei mai ei. That CL tree flower small leaves big very ugly so I not buy
‘The tree, (its) flowers are small; (its) leaves are big; (it) is very ugly: so I did not buy (it).”
Unlike Italian-type and Chinese-type languages, English-type languages, like English and French, are [-null subject] languages. Their pronouns are not permitted to be dropped, but their Infl(ection) requires that its specifier be realized overtly except
for colloquial spoken form or diary context. See the Table 1.1 for the three types of languages:
Table 1.1Three different types of languages relating to pro-drop phenomenon
Due to the different realization of pronouns and syntactic structures between [+null subject] languages and [-null subject] languages, the pro-drop phenomenon has attracted the attention of many SLA researchers. A lot of contribution to pro-drop issue has been ascribed to the second language acquisition of many European languages (Clahsen & Hong, 1995; Hilles, 1986, 1991; Lakshmanan, 1991, 1994;
Liceras, 1989; Meisel, 1991; Phinney, 1987; Prévost, 1999; Tsimpli & Roussou, 1991;
Vainikka & Young-Schilten, 1994, 1996; White, 1985); however, little attempt has been made to focus on EFL learners of Chinese (Kong, 2005; Lai, 2006; Ou, 1998;
Pan, 2008; Wen & Wu, 1997; Yuan, 1997; Zobl, 1992, 1994).
As mentioned above, the main interest in the literature about pro-drop phenomenon is about the issue of parameter resetting. It has been claimed that EFL learners of Chinese do encounter some difficulties in unlearning null arguments. The most prominent difficulty is that EFL learners tend to unlearn null subjects but not null objects in English (Chen, 1994a; Kong, 2005; Pan, 2008; Yuan, 1997; Zobl, 1992, 1994). This asymmetry brings about a problem: If EFL learners do acquire the obligatory overt subjects in English, why do they have the long-term learning problem with obligatory overt objects? This problem, first explored by Chen (1994a), is attributed to the markedness factor, in which subject positions are more emphasized than the object positions by instructors in teaching. She assumes that L1 transfer is the
Chinese-type language English-type language Italian-type language
Zero-topic √ × ×
Pro-drop √ × √
major factor in second language acquisition (SLA) unless markedness intervenes in learning. Yuan (1997) argues that AGR helps EFL learners abandon null subjects, while the L1 transfer of [+topic-drop] and the lack of positive evidence in the input data leads learners unable to unlearn null objects. He also claims that L1 transfer may first influence L2A and then parameter-resetting occurs. Against Yuan’s argument, Kong (2005), supporting Tsimpli and Roussou’s (1991) argument, suggests parameter resetting is impossible but L1-based properties would be used in L2 data. He claims that Chinese learners do not reset their parameters but adjust the L1 rule of topic chains; that is, every English sentence must have one overt topic in the sentence-initial position.
A second issue regarding the pro-drop phenomenon is whether L2 learners reset all the clustered features associated with a parameter. There are three positions on this issue. One argues that the learners would simultaneously acquire a cluster of properties related by a single parameter (Chomsky, 1982; Hilles, 1986; Haegeman &
Guéron, 1999; Hyam, 1987). Another argues that there is no obvious evidence of clustering effects, which suggests that learners seem to acquire these properties separately (Banka, 2006; Liceras, 1989). The other supports partial clustering (Hilles, 1986; Hyams, 1986, White, 1985, 1986; Cook & Newson, 1996). White (1985, 1986) indicates that L2 features may cluster together in interlanguage grammars but not entirely as predicted by the Null Subject Parameter, which specifies that syntactic and morphological properties related to a single parameter should cluster together.
However, little research has focused on the issue of clustering effects on the EFL learners of Chinese.
To sum up, in the SLA literature, the two pro-drop issues of parameter resetting and clustering effects have not been explored in depth for the second language acquisition of English by Chinese learners. Therefore, the current study aims to
investigate EFL learners’ IL in terms of the two issues related to pro-drop phenomenon. Since both L1 transfer and UG effect are considered to interplay L2 learners’ IL to a certain degree, our main interest is to examine if the parameter resetting occurs in L2A of Taiwan EFL learners. Besides, we will also examine whether pro-drop features of English cluster in Chinese interlanguage (CIL) grammars. The research questions are posed as follows for the present study:
1. Whether do EFL learners have an access to UG, like child L1 learners, or transfer their L1 knowledge in their L2 acquisition of English?
2. Do EFL learners reset their L1 value of null subject parameter to the L2 value of null subject parameter?
3. If parameter resetting is possible, do the pro-drop features of English cluster in CIL grammars?
4. If parameter resetting is not possible, what do possible reasons block the parameter resetting?