The acquisition of genitive -'sand ofby Japanese learners of English

我们对理论研究和实证研究进行了综述，包括对学习者语料库的调查和实验。在综述中，我们主要关注西山(2003年)对日语所有格no和RapPaport(2004年)的描述，以及Rozwadowska(2006年)关于英语前词所有格‘S和后名词性所有词of’的著作。在这些研究的基础上，我们描述了这三个语素的异同。在学习者语料库方面，我们使用了NICT JLE语料库(Izumi，Uchioto和Isahara，2004)，并说明了日本学习者在生成英语属格-‘S和of’方面的一些特点。我们的研究结果表明：1)日语学习者比-‘S使用得更频繁；2)他们区分这两个语素的方式与本族语者相似；3)他们产生的错误很可能是由于他们的第一语言语法；4)当他们的英语总体水平提高时，他们倾向于频繁地使用所有格；但v)在前置名词所有格标记--’S‘方面没有观察到同样的趋势。本实验的目的是考察他们对使用两个所有格标记与非动词性名词的论元有关的限制条件的知识。我们根据Haegeman和Guelon(1999)建立了我们的材料，但以英语为母语的人的行为并不像我们预期的那样，因此我们未能获得相关数据来回答我们的问题。这是因为母语人士对句子的解释和判断对语义和语用方面高度敏感，这是我们没有预料到的。我们打算在这个项目之后进行进一步的研究。

We carried out a review of theoretical works and empirical studies, including investigations into a learner corpus and conducting an experiment. In the review, we mainly focused on Nishiyama's (2003) description of Japanese genitive no and Rappaport's (2004) and Rozwadowska's (2006) works on the English pre-nominal genitive-'s and the post-nominal genitive of. Based on these studies, we described similarities and differences among these three morphemes. With regard to the learner corpus, we used the NICT JLE Corpus (Izumi, Uchimoto, and Isahara, 2004), and illustrated some characteristics in Japanese learners' production of English genitive-'s and of. Our findings showed that i) Japanese learners tend to use of more often than-'s; ii) they differentiate these two morphemes in a similar way as native speakers do; iii) they produce errors which are likely to be due to their first language grammar; iv)learners tend to frequently produce the genitive of when their general proficiency in English becomes higher; but v) the same tendency was not observed with regard to pre-nominal genitive marker-'s. We carried out an experiment with seventy-four Japanese learners of English and twenty-one native speakers of English. The purpose of this experiment was to examine their knowledge of constraints involving the use of the two genitive markers with the arguments of de-Verbal nouns. We set up our material based on Haegeman and Gueron (1999), but native speakers did not behave as we had expected, and hence we failed to obtain data relevant to answer our questions. This is because native speakers' interpretation and judgment of sentences were highly sensitive to semantic and pragmatic aspects which we did not anticipate. We intend to carry out further research after this project.