This study examined the relation between markedness and learnability as claimed in Keenan and Comrie(1997)’s Noun Phrase Accessibility Hierarchy(NPAH) in order to explicate the order of ease and difficulty in the acquisition and processing of relati...
This study examined the relation between markedness and learnability as claimed in Keenan and Comrie(1997)’s Noun Phrase Accessibility Hierarchy(NPAH) in order to explicate the order of ease and difficulty in the acquisition and processing of relative clauses(RCs) in Korean, and at the same time, to investigate some useful methods that make it easier to understand and produce Korean RC structures.
For these purposes, this study investigated the following four research questions simultaneously through the analysis of a corpus data from KSL learners as a second language and an experiment based on the corpus.
First, we examined whether the concept of markedness as claimed in NPAH is a useful tool for predicting ease and difficulty in the acquisition of Korean RCs through the analysis of a corpus from KSL and an experiment based on the corpus. Second, we attempted to determine the relation between universal typology and learnability as claimed in NPAH. In particular, we tried to determine which of the two hypotheses( Markedness generalization hypothesis(Hamilton, 1991, 1994) & Teachability‧Learnability hypothesis(Pienemann, 1984, 1998), which are considered important in SLA with respect to learnability, provides more appropriate explanation about the development stages of Korean RCs. Third, we investigated what are more effective teaching techniques in understanding and producing the structures of Korean RCs in Korean class. For this, we compared the effects between the input‐based teaching method(PI) and the meaning based output‐teaching method(MOI). Lastly, pointing out the limitations of previous studies that tried to explicate ease and difficulty in the processing and acquisition of the structures of RCs from the syntactic perspective, this study suggested that it is more reasonable not only to account for syntactic differences among the RC structures but also to include semantic perspectives in order to look at the course of processing and development of the structures of Korean RCs holistically. Then, we carried out an experiment based on our suggestion.
Above all, as a result of the analysis of the corpus of learners’ spoken and written language, it turns out that structures corresponding to RCs appear in a considerable degree only in the late part of the intermediate level. That is, the corpus analysis shows that Korean RC structures are so difficult that they are not produced in the basic level, and we think that this is because the course of utterance processing and production is affected by ‘cognitive constraints(Pienemann, 1998: 211)'.
Also, as predicted by NPAH, head nouns’ grammatical functions have influence on the production frequency of Korean RCs. That is, L2 Korean language learners produce SU RCs more frequently than DO RCs and OBL RCs. On the other hand, it is observed that there is some interrelation between head nouns’ grammatical functions and animacy. That is, learners tend to produce SU RCs in tandem with [/animate], and DO RCs and OBL RCs in tandem with [-animate]. Moreover, the typological output patterns of RC structures in learners’ corpus are also found in the results of the analysis of corpuses like native Korean speakers’ corpus, textbook corpus, and corpus from learners other than those who have Chinese as their mother tongue. That is, SU RCs are used about twice as frequently as DO and OBL RCs in the comparative corpus. In addition, as is the case with the tendency in learners’ corpus, ‘the one‐to‐one principle(Anderson 1984; re‐cited from Ellis 1994: 380-381)' is also found between specific types of RCs and encoding of animacy in the comparative corpus. That is, like native Chinese speakers, native Korean speakers and L2 Korean language learners prefer encoding SU RCs with [/animate], and DO and OBL RCs with [-animate]. Thus, it is hard to decide from the results of our analysis of learners’ corpus whether head nouns’ syntactic functions determine animacy or vice versa. Consequently, in order to determine ease and difficulty in the acquisition and processing of Korean RC structures, we should not only look at head nouns’ syntactic relations as claimed by NPAH but also consider together the frequency of inputs from ambient languages, transfer from learners’ mother tongue, and above all, head nouns’ semantic features(animacy). Thus, reflecting the results of this study’s corpus analysis that animacy can have influence on the grammatical system of RCs, we presented some experimental questions by controlling head nouns’ grammatical functions and the features related to animacy. Thirty four basic learners whose mother tongue is Chinese participated in the experiment, and their knowledge of RC was measured through SCT(sentence combination production tasks) and PST(picture selection tasks).
The results show, first, that with respect to NPAH’s animacy, ease and difficulty in the acquisition and processing of each Korean RC structure partly agree (SCT: DO=SU>OBL) and partly does not agree with the NPAH hypothesis (PST: OBL>DO=SU). This means that ease and difficulty in the acquisition and processing of Korean RC structures can be explained not by the simple relations with head nouns but by difference in word order between L1 and L2 as well as in animacy. That is, it is likely that the reason why the structures of DO RCs in production tasks show higher average accuracy than those of SU RCs and OBL RCs is that an independent processing step requiring some cognitive load is unnecessary when a head noun has the grammatical function of DO in a Korean RC structure as the word order inside the RC (Korean: [NP V] Head) becomes identical to that in the learner’s mother tongue (Chinese: [NP V] Head). On the other hand, in case of Korean SU RCs, it is because the learner should move some of the constituents independently in accordance with the word order of the target language due to difference in word order among the constituents within the RC (Korean: [NP V] Head; Chinese: [V NP] Head). That is, as for the ease and difficulty of RC structures, we should consider the word order among the constituents of the modifying clause as well as the ‘basic word orders’ of L1 and L2 (Hakuta, 1981).
Second, with respect to the possibility of generalization in learning, though they received an instruction focused on the structure of DO RC, a relatively unmarked structure, the learners turned out to generalize their knowledge of RCs to other structures that they had not been taught. Here, the generalization is not mono‐directional but bi-directional differently from the claim of the Zobl(1993, 1985)'s Projection Model.
Third, as for the effect of teaching on RCs, formal classroom education is effective in increasing the rate of acquisition of Korean RC structures. However, the results of comparison between PI- input‐based teaching method-, and MOI- meaning based output teaching method show that there are no statistically significant differences between the two groups. For one thing, however, the results from the technical statistic analysis show that the PI group is more effective than the MOI group in both understanding tasks and production tasks. Especially, given that the PI learners who participated only in input activities were different from the learners in the MOI group who took part in meaningful output activities, we found that PI is a very effective teaching technique not only in understanding target structures but also in producing them. These results also constitute evidence showing that enhanced input data, explicit meta‐linguistic information, and structured input are very effective in increasing the rate of L2 learners’ acquisition of Korean RCs. However, the teaching effects of the PI and MOI techniques do not appear to have influence on the stage of complete reconstruction with the learners’ interlanguage system.
Reflecting the results from this study, in teaching‐learning scenes with RCs, we need to consider the syntactic functions of RCs with respect to head nouns, animacy, word order, and the phenomenon of transfer from L1 as much as adnominal endings, and especially it is necessary to make teaching materials and syllabuses by reflecting learners’ strategies for processing RCs in accordance with task variables.
，韩语论文题目，韩语毕业论文 |
