Acquisition of Chinese relative clauses

Chang (1984) investigated the comprehension of Chinese relative clauses by children aged 1, 2, 4, and 6. Four types of relative clauses were involved: SS, SO, OO, OS (Table 2.11). Data were elicited by asking subjects to act out the meaning of testing sentences. Three hypotheses of relative clause were tested: parallel function hypothesis (PFH) (Sheldon, 1977), N-V-N strategy (Bever, 1970), and conjoined clause hypothesis (Tavakolian, 1981). Recall that PFH predicts the order of difficulty: SS, OO>SO, OS, based on the notion that the head noun having identical functions both in the main and relative clauses makes the comprehension of relative clause easier. N-V-N strategy upholds the idea that sentences in canonical N-V-N order are less demanding than non-canonical order ones. This hypothesis assumes that when children have difficulty understanding a complex sentence, they tend to parse them in the N-V-N order. Thus, applying this notion in relative clauses, it is hypothesized that children would perform well on OS sentences, but poorly on SO in English. Conjoined clause hypothesis is based on the assumption that children interpret the first NP as the subject of both verbs when comprehending complex sentences. In the case of relative clauses, children will find SS easy to process and OS hard to process. However, this hypothesis has difficulty in predicting children’s

performance on OO and SO because these two constructions are not in the normal English order of SVO. The results demonstrated that children were not able to fully comprehend relative clauses even at the age of 9. Generally, the order of difficulty of Chinese relative clause was found to be SS=SO>OS=OO (>means is easier than), with which Chang claimed that none of the orders predicted by the previously discussed hypotheses matched. Finally, the author reached the conclusion that the order of difficulty of Chinese relative clause can be best accounted for by the

interruption in the main sentence. To be specific, in Chinese, OS and OO are center embedding, the interruption of which in the matrix sentences makes them harder to process than SS and SO, the left embedding relative clause types.

Table 2.11 Four basic constructions of relative clause (from Chang, 1984) Function of head noun in

the main sentence

Function of head noun in relative clause

Example

Subject (S) Subject (S) 咬狗的貓追老鼠

Subject (S) Object (O) 被狗咬的貓追老鼠

Object (O) Subject (S) 老鼠追咬狗的貓

Object (O) Object (O) 老鼠追被狗咬的貓

Table 2.12 The order of difficulty as predicted by the three hypotheses Hypothesis Order of difficulty in English

(>means is easier than)

PFH SS, OO>SO, OS

N-V-N strategy OS>SS, OO>SO Conjoined clause hypothesis SS>OS

The same order of difficulty of Chinese relative clauses was obtained in the study by Cheng (1995). As with Cha ng, one of the intents of Chengs’ study was to see the order of difficulty of the four types of Chinese relative clause, SS, SO, OS, OO, in young children’s interlanguage. A measuring task similar to that in Chang’s

was employed: the act out task, in which the subjects were asked to act out the meaning of the sentences. A second task was meant to elicit subject’s relative clauses: a picture-telling task, in which a pair of pictures with minimal contrast was involved. In the production task, it was found that even the youngest group was able to produce relative clause successively. The results showed that children appeared to have acquired relative clauses at an early age, in contrast to what has been revealed in Chang’s study. In the productive data, it was also noted that many of the subjects employed the gap strategy to form Chinese relative clauses while the few others formed relative clauses with the resumptive pronoun and resumptive NP occurring.

Chiu (1995) probed ten types of Chinese relative clause in Chinese-speaking elementary school students. The ten types of relative clauses were SS, SO, SIO, SP, SG, OS, OO, OIO, OP, OG, where the first letter represents head noun function in the matrix sentence whereas the second one stand for head noun function within the relative clause. From the productive data, it revealed that subjects produced more headless relative clauses than headed ones in general. The number of headless relative clauses increased as the age increased. Three different strategies were employed to form relative clauses: gap, RP (resumptive pronoun), and RNP (resumptive NP). As for the comprehension test, the author adopted the same measuring instrument as in Chang’s and Cheng’s studies, the act out task. The results of the comp rehension test replicated those of Chang’s and Cheng’s studies: the relative clauses in the subject matrix positions posed less difficulty for the subjects across all age groups to comprehend than did the relative clauses in the object matrix positions. The order of difficulty therefore were SS, SO, SIO, SP, SG>OS, OO, OIO, OP, OG, (> means are easier than) for which the author suggested that the difficulty in comprehending the center-embedding relative clauses that interrupt the processing flow might be the leading cause.

While the above three studies on the Chinese relative clauses suggest that the object matrix sentences are more difficult than subject matrix sentences on the ground that center-embedding causes interruption and thus the difficulty, all of the three studies do not differentiate the difficulty level among the relative pronoun positions.

According to the structural Distance Hypothesis and Hamilton’s SOHH, Shyu (2005) argues that subject position is easier than object position. Therefore, the difficulty order of the four types of Chinese relative clauses would be OS > SS > OO

> SO.

Table 2.13 The order of difficulty of Chinese relative clauses in the four studies Study Main finding Order of difficulty (> means is easier than) Chang (1984) Center-embedding

can best account for the difficulty of relative clause

formation in Chinese.

SS=SO>OS=OO

2.4.4 Summary

In the Chinese relative clause acquisition, we can see from Table 2.13 that the subject relative clause is the easiest one for Chinese children and the most frequently used type of relative clause as the various corpora show. However, in terms of the matrix clause, there seems no consistent order: while the children perform best in subject matrix clause, the corpora show that the object matrix clause occur s most frequenctly.

From the review, as most researchers have acknowledged, the universal markedness as predicted by NPAH plays a critical role in shaping second language learners’ relative clause acquisition. However, it still can not be denied that native language exert influence to some extent and at some point of time in the learning process. And examining learners’ way of avoiding to produce relative clauses seems to be a promising means in which we actually see how transfer works. In addition, it can be also seen that center-embedding that cause the interruption of processing flow in the main clause is of crucial importance in the formation of relative clauses.

However, the focus seems to have been limited on the four types of relative clauses:

SS, SO, OO, SO. This clearly is a limitation and further research is required in order to make more contribution to our understanding of second language acquisition.

CHAPTER THREE MEHTODOLOGY

This chapter first presents the research questions for the thesis. Then, it specifies the subjects of the study, the research design, the types of relative clause under investigation, and the adoption of the three measuring tasks: the sentence combination task, the Chinese-English translation task, and the grammaticality judgment task. Last is the delineation of the procedure of data collection, and the scoring method.

3.1 Research questions

In view of the literature review, we have the following research questions for the present study:

1. Does Taiwan EFL learner’s acquisition of English relative clauses follow the order of acquisitio n as predicted by NPAH?

2. Is the universal hierarchy as predicted by NPAH for the resumptive pronouns retained in the Taiwan EFL learners’ interlanguage of relative clauses?

3. Are relative clauses that are center-embedded more difficult than right-embedded as is predicted by PDH in Taiwan EFL learners’ acquisition of English relative clauses?

4. What’s the role of sentence discontinuity in the main and relative clauses, as predicted by SO Hierarchy, in Taiwan EFL learners’ acquisition of English relative clauses?

5. What are the avoidance strategies that will be demonstrated in Taiwan EFL learners’ acquisition of English relative clauses and what’s the tendenc y going to be?