Processing Subject and Object Relative Clauses in Japanese by Turkish Speaking Learners Barış KAHRAMAN Çanakkale Onsekiz Mart University Keywords: Sentence processing, subject relatives, object relatives, Turkish learners of Japanese 1 Introduction It is well known that there is a processing asymmetry between subject and object relative clauses (SRs and ORs, respectively) in first and second languages (L1 & L2). In L1 sentence processing of Japanese, previous studies have consistently reported that SRs are easier to process than ORs (e.g., Ueno & Garnsey, 2008). In the case of L2 Japanese, previous studies have reported different results. For example Kashiwagi (2011) showed that English speaking learners of Japanese (ELJ) read ORs faster than SRs. On the other hand Mitsugi et al. (2010) showed that Korean speaking learners of Japanese (KLJ) read SRs faster than ORs as in the case of Japanese native speakers, whereas ELJ processed SRs and ORs in the same pace. Mitsugi et al. (2010) argued that learners’ L1 might have influenced the processing patterns of Japanese learners. If learners’ L1 has an impact on the L2 sentence processing as argued by Mitsugi et al. (2010), the learners whose L1 is similar to Japanese may display similar processing pattern as KLJ and Japanese native speakers (JNS). At the present time, the number of RC processing studies in L2 Japanese and learners’ L1 background is very limited. In order to explore the possible effects of linguistic similarity between L1 and L2 on the sentence processing of Japanese, we need to investigate various L1 speaking learners’ processing pattern. Turkish speaking learners of Japanese (TLJ) provide a good test case in Japanese, because Turkish is an SOV language, and word order of relative clauses (RCs) is identical to Japanese. Moreover case markers are attached to nouns within RCs as in Japanese and Korean. Therefore, it can be expected that TLJ may also process Japanese RCs as KLJ and JNS do. The current study aims to explore how Turkish speaking learners process Japanese RCs, and test whether the learners process L2 sentences as native speakers do when their L1 is similar to L2. In order to answer this question a self-paced reading experiment was conducted. 2 Experiment Twenty-six higher-intermediate students who study Japanese as their major at Çanakkale Onsekiz Mart University participated in the experiment. They were all undergraduate students (3rd or 4th year) and native speaker of Turkish. Experimental sentences consist of 24 sets of SRs and ORs and 72 filler sentences. The experimental sentences were divided into two lists by a Latin Square design, and each participant was assigned to one list. Experimental sentences were randomly presented on the computer screen in word-by-word, non-cumulative self-paced reading task by Linger 2.94 (Douglas Rohde). After reading a sentence, participants answered a yes-no question. The semantic plausibility of test sentences was confirmed in a norming study.
Example of experimental sentences: (1a) SR: Niwa-de otokonoko-o oshita onnanoko-wa tanoshiku Garden-LOC boy-ACC pushed girl-TOP funnily ‘The girl who pushed the boy in the garden was playing funnily.’ (1b) OR: Niwa-de otokonoko-ga oshita onnanoko-wa tanoshiku Garden-LOC boy-NOM pushed girl-TOP funnily ‘The girl who the boy pushed in the garden was playing funnily.’
asonde-i-ta. play-PRG-PST asonde-i-ta. play-PRG-PST
3 Results and Discussion The accuracy rate of comprehension questions was 96% in SRs and 94% in ORs. This difference was not statistically significant. The results of ANOVAs for repeated measures showed that the reading times of the head noun (onnnanoko-wa) in SRs and ORs differed significantly [F1 (1,25) = 7.74, p < .01; F2 (1,23) = 3.88, p = .06]. In other regions, there was no significant difference.
The results show that TLJ read the head noun of SRs 181 millisecond faster than that of ORs. This result is consistent with Mitsugi et al. (2010) and previous L1 studies in Japanese (e.g., Ueno & Garnsey, 2008), indicating that Turkish speaking learners of Japanese processed Japanese SRs more easily than ORs, as Korean speaking learners and native speakers of Japanese do. On the other hand, TLJ’s processing pattern obviously differs from English speaking learners. Taken together with previous studies, the present results suggest that when the learners’ L1 is structurally similar to Japanese, they process Japanese sentences in a similar way to native speakers (Mitsugi et al., 2010). Moreover, the results of current study imply that the structural similarity between L1 and L2 might be an important factor in native-like L2 sentence processing (Juffs, 1998). References Juffs, A. (1998) Main verb versus reduced relative clause ambiguity resolution in L2 sentence processing. Language Learning, 48, 107-147. Kashiwagi, A. (2011) Processing relative clauses in first and second language: A case study. Unpublished PhD Dissertation. The Ohio State University, OH. Mitsugi, S., MacWhinney, B. and Shirai, Y. (2010) Cue-based processing of relative clauses in L2 Japanese. Selected Proceedings of the 2008 Second Language Research Forum, in Matthew T. Prior et al. (eds.), 123-138. Somerville, MA: Cascadilla Proceedings Project. Ueno, M., and Garnsey, S. (2008) An ERP study of subject and object relative clauses in Japanese. Language and Cognitive Process, 23, 646-688.
