In general, as the result showed, the participants’ acquisition of English relative clauses was constrained by the universal principle, except in the grammaticality judgment test, which turned out to be surprisingly unusual in the result. In this test, the learners seemed to experience great difficulty in the typologically least marked position, SU, rather than the typological marked ones, which totally defied the universal hierarchy by NPAH. Nor was there any support for PDH and SOHH.
Furthermore, the pattern in which learners retained the resumptive pronoun was not in
the order as predicted by NPAH. Upon close examination, it revealed that some of the items for the subject relatives were where the problem was, as in item 2, replicated below. Research has discovered that Chinese learners tended to process sentences like item 2 as a topic-comment construction rather than as relative clause construction;
thus they didn’t consider this construction wrong for lacking relative makers (Yip, 1995; Gass and Selinker, 2001). That is, these learners were following an
appropriate native language pattern of establishing a topic and then making a
comment about it. This is not unlike the following construction in English: You see the man? He just ran a red light. as opposed to Did you see the man who just ran a red light? In addition, the participants seemed to have much trouble in locating the error in items 12 and 14, which turned out to receive the lowest scores, thus making the SU to be the most difficult relatives to the learners in this study. At first sight, it seemed that the length of item 14 may be the cause of the difficulty, but this was not conclusive yet. To clarify the problem, a separate study is called for.
Item 2: *In our village, there were many people didn’t have much money. (OS) Item 12: *Almost all of the people appear on television wear makeup. (SS) Item 14:* People who work in the hunger program they estimate that 3500 people in the world die from starvation every day of the year. (SS)
5.3 GEN, IO and OPREP relatives
Although the accuracy order for the relative clauses largely complied with the three predictor hypotheses, minor discrepancy existed, namely in the order for GEN and that between IO and OPREP. In the IO and OPREP part, it was long indicated by Keenan (1975) that “IO and OPREP were indistinguishable in most languages, including English, as both were expressed as “prepositional phrases”. Empirically,
the indistinguishable relationship between IO and OPREP was confirmed by a number of studies, namely Hyltenstam (1990), Pavesi (1986), and Tarallo and Myhill (1983), as summarized in Table 5.1. Hyltenstam (1990) has clearly pointed out that
“the reverse order for OPREP and IO might be suggestive of an order that is more likely to reflect the universal psychological reality of learning relative clauses”.
Accordingly, some researchers simply put the two relative pronouns into one category as they devised their experiment (e.g., Izumi, 2003; Gass, 1979)
In the order for GEN, the present study has the same research result as in Gass’s (1979, 1980) study. Gass made the speculation that the high accuracy rate for the GEN relatives may be attributable to the uniqueness and saliency of this function.
This however was not conclusive. Besides, while the GEN used in this study was subject function, Gass didn’t mention this. While not refuting the speculations made by Gass, there are other possibilities that might serve to explain the high rank of GEN in the present study. First, the GEN of subject function has a canonical word order as SU does, so learners may have produced GEN in the same way they produced SU, thus resulting in the high accuracy rate for GEN in the present study. Furthermore, according to the Genitive Hierarchy by Jones (1991, in Ellis, 2003) and the processing discontinuity contended by Hamilton (1994, for detailed account please refer to Chapter one), GEN of subject function has only one processing discontinuity just like SU. The structural ease of processing may further account for the high rank of GEN in this study.
The result of retention rate for GEN seems to support the assumptions I made.
That is, learners rarely supplied resumptive pronoun for GEN relative clause in the test. According to NPAH, the GEN is a relatively marked form as it takes the second lowest position in the hierarchy. And the retention of the pronoun serves to ease the
However, the GEN relatives did not actually pose a processing problem when the learners were forming the relative clauses; so there was no need for the learners to provide a resumptive pronoun for to ease the processing difficulty.
Table 5.1 The accuracy order in previous studies
Study Order of accuracy
Gass (1979, 1980)
SU/GEN>DO>IO>OPREP>OCOMP Hyltenstam (1990) SU>DO>OPREP>IO>OCOMP>GEN Pavesi (1986) SU>DO>OPREP>IO>OCOMP>GEN Tarallo and Myhill (1983) SU/DO>OPREP>IO>GEN
5.4 The interference of Chinese
Transfer of L1 characteristics has been invariant ly indicated in a lot of research (Tarallo and Myhill, 1983; Gass, 1979, 1980; Hyltenstam, 1990; Pavesi, 1986). This study also manifests the influence of the learners’ native language in several ways.
Overall, the data in the present study support the hypothesis that
center-embedded relative clause is more difficult than right embedded one as predicted by PDH. However, the Chinese- English translation test in this study ran counter to PDH, where the right-embedded sentences received slightly higher scores than center-embedded sentences. In this case, the influence of Chinese language was suggested on the ground of the divergent sentence configuration between Chinese and English in relative clause. Contrary to English relative clause, which has main clause interruption in subject matrix positions, Chinese relative clause has the interruption in object matrix positions, the result of different branching direction (Schachter, 1974; Flynn, 1989a, 1989b). Consequently, the sentence discontinuity in
object matrix sentence makes the Chinese reading harder to comprehend than subject matrix sentence (Chang, 1984; Cheng, 1995; Chiu, 1995), thus affecting the learners’
ability to translate the Chinese relative clauses placed in the object matrix positioning into English. There was the case that the learners explicitly expressed the difficulty they were experiencing in comprehending the object matrix positioning sentences in Chinese: the learners made remarks on the test sheet, pointing out that the object matrix sentences were hard to understand. In addition to the empirically tested hypothesis that the structure of center embedding of Chinese relative clause increases the comprehension burden, there is the research finding that Chinese learners rarely use relative clauses in real life Chinese may further account for the learners’ trouble in interpreting Chinese relative clauses (Houng, 1986).
The analysis of the learners’ avoidance behavior further revealed the effect of Chinese in the learners’ acquisition of relative clause. For example, quite a few subjects produced the English translation for item 5, an IO relative in this study, as the one illustrated below, which was apparently produced in the order of Chinese reading.
It was obvious that the learners had trouble comprehending the semantic meaning of the Chinese reading and made such errors.
Item 5: 他認識教授寄了一封信給她的那一個女孩。
→The professor that he knows sent a letter to the girl.
Besides, the evidence of potential interference from Chinese was evidenced in the total number of avoidance. Where it was 175 cases of avoidance for the sentence combination test, it was 219 ones for Chinese-English translation test.
The effect of Chinese influence can also be seen in the tendency of pronoun retention in the Chinese- English translation test. That is, the subjects in this study
tended to maintain resumptive pronoun in IO and OPREP relatives more often than in OCOMP relatives. Since the pronoun retention in Chinese IO and OPREP relatives is obligatory and that Chinese does not have OCOMP relativization, a possible explanation for this tendency may be that the subjects transferred the Chinese feature (i.e., the supply of resumptive pronoun) in the formation of English relative clauses.
Ellis (2003: 425) has had the following conclusion for Gass’s (1979, 1989) and Hyltenstam’s (1990) studies:
The conclusion was that markedness interacts with L1 transfer in the sense that although pronominal copy errors were more likely in the lower positions and less likely in the higher positions of the (NP)AH, the overall error frequency reflected whether or not pronominal copies occurred in the learners’ L1.
5.5 The implication of UG
Based on the argument of the poverty of the stimulus, Chomsky (1965) resorted to the concept of Universal Grammar. The principle argument was that given the degenerate input, spoken language that is full of false starts, slips of the tongue, etc., children still create a mental representation of language which is not only much more complex than could be expected, but also strikingly similar to that of other native speakers of the same language. That is, input alone can not explain the L1 acquisition, and, therefore, the biological endowed Universal Grammar must exist and make the task facing the child much easier, by equipping them in advance with a clear set of expectations (i.e., principles and parameters) about the shape of the language they are to learn.
If this argument is true about L1 acquisition, it is also true about second language
acquisition (Cook, 1988; Flynn, 1987). Cook (1988) claims that, like L1 learners, L2 learners possess knowledge of the L2 that they could not have acquired from the input and which must have existed within their own minds. Flynn (1987) believes that adult second language learners have continued access to UG just like L1 language learners. In addition, the critical role of L1 is acknowledged by Flynn (1989a, 1989b). She argues that in cases where the L1 and L2 parameter settings are the same, learning is facilitated because the L2 learners are able to consult the structural configuration established for the L1 in the construction of the L2 grammar. Where the L1 and L2 parameter settings are different, the learner has to assign new values and although this is not problematic, it does add to the learning burden.
In the case of this study, in which the subjects are Taiwanese foreign language learners, the logical problem of language acquisition is just much the same as L2 language acquisition. The language input, which comes mainly from school
instruction, is scarce. However, as the result shows, there is still up to 90% accuracy rate (in the case of OGEN type of sentence in the sentence combination test). Even in the acquisition of OCOMP relative clause, which function Chinese does not
manifest, and the ESL textbooks here in Taiwan do not have entry for, there is still up to 40% accuracy rate. In view of this, although this study is not UG based, the theoretical reasoning of UG can still apply and accordingly the implication may follow: UG may still be operative in the minds of the Chinese adult foreign learners, otherwise the task of acquiring English relative clause, a head-initial construction as opposed to the head-final construction of Chinese relative clause, would be
impossible in view of the extremely scant input the subjects receive in the context.
5.6 Summary and Conclusion
This study set out to examine the Taiwan college students’ underlying knowledge of English relative clauses in order to see what the factor(s) is/are that constrain(s) the learners’ acquisition process, whether it be the universal factor, the native language, or human general problem solving skill. Three predictor hypotheses were used in this investigation: NAPH, PDH, and SOHH, which are motivated by different theoretical backgrounds. The NPAH is based on the typological markedness, the PDH is based on the notion of human limited capacity of short term memory, and SOHH the
combination of NPAH and PDH, and the structural difficulty of relative clause.
Three kinds of tasks were used to elicit the data: sentence combination test,
Chinese-English translation test, and grammaticality judgment test. According to the research result, the main findings of this study are summarized as follows:
1. The learners’ acquisition of relative clauses is largely constrained by the universal markedness by NPAH, except GEN, and the order between IO and OPREP.
2. The retention of pronoun is largely constrained by the linguistic universals of NPAH, but Chinese also has certain influence in the learners’ choice of supplying resumptive pronoun.
3. The learners largely did experience more difficulty in center-embedded relative clauses, which matches PDH.
4. Largely, SOHH is a valid prediction of the Learners’ acquisition of relative clauses.
5. The learners tended to avoid on the positions low on the NPAH, except GEN.
6. The learners did frequently avoid OCOMP relatives and center-embedded relative clauses.
Although the outcome in grammaticality judgment task does not have any support to the three hypotheses due to the exceptional nature of some of the test items, the overall outcome suggests that NPAH, PDH, and SOHH are in unification relationship, together making their contribution to our understanding of the acquisition problem of relative clauses. That is, the fact that the data in the other two tasks (except for the opposite result for PDH in the translation test) partially support the universal
markedness as predicted by NPAH, but fully support PDH and almost fully support SOHH may imply that learners are acquiring the relative clauses through the path that is constrained by the linguistic universals of NPAH and the universal of human cognition as hypothesized by PDH.
The framework of universal grammar has been proposed by Chomsky (1965) to account for the logical problem of child first language acquisition, and thereafter been applied in the adult second language acquisition (Flynn, 1987, 1989a, 1989b; White, 1996, 2003). It is different from the typological study of linguistic universal that was proposed by Greenburg (1966) in the methodological terms, but the pursuit of commonalities of the world languages is not dissimilar (Ellis, 1996c; Ritchie and Bhatia, 2003). Hence, this study’s outcome that the typological markedness is largely complied with may suggest that the existence of UG be operative in the adult foreign language learners. The fact that the learners are able to acquire OCOMP relative clause can also be accounted for in this framework.
While it is acknowledged that the linguistic universal factor and human cognitive mechanism are at play in Chinese adult learners’ language acquisition, the role of L1
strategies the learners made in the formation of English relative clauses. However, the influence of Chinese does not make the task of language learning totally
impossible since the learners are able to reset the parameter to that of the target language by assigning new values to the existing L1 parameter setting (Flynn, 1987).
Yip (1995) has made the following statement made:
It is proposed that from the erroneous constructions learners produce, we shall see that interlanguage is the product of a complex interaction between the native and target grammars and universal principles underlying grammar construction.
This thesis has probed a number of issues that have never been addressed in the Taiwan EFL situation. However, the results are not conclusive yet unless more relevant research is to be done in similar context. The results obtained for the GEN in this study is interesting in a way that was hardly mentioned before (except Gass, 1979, 1980). It is thus hoped that GEN relative clause can be looked into more closely and extensively in order to help us better understand language acquisition.
The fact that the participants in the study were able to form OCOMP relative clause (which does not manifest in Chinese and is hardly grammatical in the eyes of the native speakers of English) is an intriguing issue as well. It is a question that calls for further stud ies whether it is the universal principles that are at work or not.
REFERENCES
Azar, B.S. (1999). Understanding and Using English Grammar. New York: Pearson Education.
Bever, T. G. (1970). The cognitive basis for linguistic structures. In J. R. Hayes (ed.) Cognition and The Development of Language. New York: Wiley.
Braidi, S. M. (1999). The Acquisition of Second Language Syntax. London: Arnold.
Celce-Murcia, M & Larsen-Freeman, D. (1999). The Grammar Book. USA: Heinle &
Heinle Publishers.
Chang, H. W. (1984). The comprehension of complex sentences by children: relative clause. Chinese Journal of Psychology, 26, 57-66.
Chen, J. J. (2004). A Study on Relative Clauses in Taiwanese EFL Students’ Writing.
M.A. Thesis, National Kaohsiung Normal University.
Chen, X. L. (2004). An Analysis of College Students’ Knowledge of English Relative Clauses. Unpublished Paper, NSYSU.
Cheng, S. (1995). The Acquisition of Relative Clauses in Chinese. M.A. Thesis, National Taiwan Normal University.
Chiu, B. (1995). The Nature of Relative Clauses in Chinese-Speaking Children.
National Science Council Research Report, NH NSC85-2418-H003-001.
Chomsky, N. (1965). Aspects of The Theory of Syntax. Cambridge: MIT Press.
Comrie (1989). Language Universals and Linguistic Typology. Oxford: Basil Blakwell Ltd.
Comrie, B. (1989). Language Universals and Linguistic Typology. Oxford: Billing &
Sons Ltd, Worcester
Cook, V. (1988). Chomsky’s Universal Grammar: An Introduction. Oxford: Basil Blackwell.
Cook, V. (1993). Linguistics and Second Language Acquistion. New York: St.
Martin’s Press.
Cook, V. (1996). Second Language Learning And Language Teaching. New York:
Arnold.
Doughty, C. (1991). Second language acquisition instruction does make a difference:
evidence from an empirical study of SL relativazation. Studies in Second Language Acquisition, 13, 431-469.
Dulay, H., and Burt, M. (1973). Should we teach children syntax? Language Learning 24, 245-258.
Eckman, F. R., Bell, L., & Nelson, D. (1988). On the generalization of relative clause instruction in the acquisition English as a second language. Applied
Linguistics, 9(1), 1-20.
Ellis, R. (1996a). The role of the first language. In Ellis, R. (ed.), Understanding Second Language Acquisition (pp.19-41). Oxford: Oxford University Press.
Ellis, R. (1996b). Interlanguage and the “natural” route of development. In Ellis, R.
(ed.), Understanding Second Language Acquisition (pp.42-74). Oxford:
Oxford University Press.
Ellis, R. (1996c). The universal hypothesis and second language acquisition. In Ellis, R. (ed.), Understanding Second Language Acquisition (pp.190-214). Oxford:
Oxford University Press.
Ellis, R. (2003). The study of Second Language Acquisition. Oxford: Oxford University Press.
Ellis, R. (2005). Measuring implicit and explicit knowledge of a second language: a psychometric study. SSLA, 27, 141-172.
Flanigan, B. O. (1995). Anaphora and relativization in child second language acquisition. SSLA, 17, 331-351.
Flynn, S. (1987). A Parameter-setting model of L2 acquisition: Dordrech: Reidel.
Flynn, S. (1989a). The role of the head-initial/head- final parameter in the acquisition of English relative clauses by adult Spanish and Japanese speakers. In Gass, S.
M. and Schacher, J. (eds.). Linguistic Perspectives on Second Language Acquisition (pp.89-108). Cambridge: Cambridge University Press.
Flynn, S. (1989b). Spanish, Japanese and Chinese speakers’ acquisition of English relative clauses: new evidence for the head-direction parameter. In Hyltenstam, K. & Obler, L. K. (eds.). Bilingualism Across The Lifespan: Aspects of
Acquisition, Maturity, And Loss (pp.116-131). Cambridge: Cambridge University Press.
Gass, S. (1979). Language transfer and universal grammatical relations. Language Learning, 29(2), 327-344.
Gass, S. (1980). An investigation of syntactic transfer in adult second language learners. In Scarcella, R. & Krashen, S. (Eds.), Research in Second Language Acquisition: Selected Papers from the Los Angeles Second Language
Acquisition Research Forum (pp.132-141). Rowley, MA: Newbury House.
Gass, S. (1996). Second language acquisition and linguistic theory: the role of language transfer. In William C. Ritchie and Tej K. Bhatia (Eds.), Handbook of second language acquisition (pp. 317- ). San Diego: Academic Press Gass, S. M. & Selinker, L. (2001). Second Language Acquisition: An Introductory
Course. New Jersey: Lawrence Erlbaum Associates.
Greenburg, J. H. (1966). Universals of Language. New York: The M.I.T. Press.
Hamilton, R. L. (1994). Is implicational generalization unidirectional and maximal?
Language Learning, 44, 123-157.
Houng, C. C. L. H. (1986). The Relative Clause Construction In Chinese : Noun
Austin.
Hyltenstam, K. (1987). Markedness, language universals, language typology, and second language acquisition (pp, 55-78). In Pfaff, C. (Ed.), First and Second Language Acquisition Processes. Boston: Heinle & Heinle Publishers.
Hyltenstam, K. (1990). Typological markedness as a research tool in the study of second language acquisition. In Dechert, H. (ed.). Current trends in European
Hyltenstam, K. (1990). Typological markedness as a research tool in the study of second language acquisition. In Dechert, H. (ed.). Current trends in European