The Acquisition of Point of View and Voice in Japanese as a Foreign/Second Language : The influence of the linguistic and non-linguistic environment

本研究考察了日语作为外语/第二语言（JFL/JSL）学习者日语语音系统的习得情况，包括直接被动语态（DP）、间接被动语态（IP）和受益语态；在日语中，它们都是表示转变或观点的A。Tanaka早期的研究表明，在复杂句/并列句中，母语人士使用被动语态或助人为乐来统一观点。然而，非母语人士却做不到这一点。数据来源于生产测试，使用图片提示从被试中引出上述结构和复合句。本研究采用了两种方法：(1)对英语教学学习者的语言进行横断面研究；(2)对JSL学习者进行随访研究。结果表明，在JFL和JSL环境中，生成有益句和复合句的能力都有所提高。但是，在JSL环境中使用shiten和IP的能力略有提高，而在JFL环境中则没有。此外，在助动词方面，与主动语态具有相同观点的“te-kureru”比与被动语态具有相同观点的“te-morau u”产生得更早。对于IP，学习者倾向于使用DP结构。最后，为了明确日语语音系统的习得顺序，我们对英语使用者在JSL和JFL语境下的数据，以及中文、印尼语、马来语和韩语使用者在JSL语境下的数据使用了隐含缩放。结果表明，无论上下文或L1如何，首先产生的是受益人，然后是DP，最后是IP。这三个类别形成了一个隐含的层次结构。从对韩语、汉语和英语国家安全局的口语熟练程度访谈中获得的生产数据支持了日语语音系统术语的收购订单。

This study examired the acquisition of the Japanese voice system by learners of Japanese as a foreign/second language (JFL/JSL), including the direct passive (DP), indirect passive (IP) and the benefactive ; all function as A means to signal shiten or point of view in Japanese. Earlier studies by Tanaka have indica tednative speakers use the passive voice or the benefactive to unify the point of view in complex/compound sentences. However, non-native speakers fail to do so.Data were obtained froma production test which elicited the structurse mentioned above and complex/compound sentences from subjects using picture prompts. Twoapproaches were used : (1) a cross-sectional study of TEL learner language ; and (2) a follow-up study with JSL learners.The results indicate that the ability to produce the benefactives and complex/compound sentences improves inboth JFL and JSL environments. However, the ability to use shiten and the IP improves slightly in the JSL context, but not in the JFL environment. Moreover, concerning the benefactive, 'te-kureru, 'which has the same point of view as the active voice, is produced at an earlier stage than te-morau u, which has the same point of view of the passive voice. Concerning the IP, learners tend to use the DP structure.Finally, in order to clarify the acquisition order for the Japanese voice system, implicational scaling was used for the data of English speakers in both JSL and JFL contexts, and Chinese, Indonesian, , Malay, and Korean speakers in JSL contexts. The results indicate that, irrespective of the context or the L1, the benefactive is produced first, then the DP, and finally the IP.These three categories form animplicational hierarchy. Prodution data from oral proficiency interviews with NSs of Korean, Chinese, and English support an acquisition order for the Japanese voice sys tern.