@article{oai:repository.ninjal.ac.jp:00002196,
 author = {前川, 喜久雄 and MAEKAWA, Kikuo},
 journal = {日本語科学, Japanese Linguistics},
 month = {Oct},
 note = {国立国語研究所, The National Institute for Japanese Language, 本稿の前半では筆者らが現在構築を進めている『現代日本語書き言葉均衡コーパス』(BCCWJ)の概要と特徴を紹介し,後半ではBCCWJやそれをさらに発展させた大規模均衡コーパスが言語研究にどのような影響を及ぼすかについての予測を述べた。類義語の研究やコロケーションの研究のように,従来から行われてきた研究がコーパスによって一層進展すると期待されるcorpus-basedな研究のほかに,コーパスなくしては行いえないcorpus-drivenな研究も考えられる。その一例として,文法性判断の個人ないし状況による異同について考察し,文法性判断は言語刺激との接触経験によって容易に影響を蒙ることを指摘した。最後に文法性判断の異同をコーパスによって説明するためには最低でも数十億語規模のコーパスが必要になることを指摘した。, The aim of this paper is twofold. In the first half of the paper, design issues of the Balanced Corpus of Contemporary Written Japanese were discussed paying special attention to the recent criticisms against the manual (as opposed to 'automatic') construction of a large-scale balanced corpus. The last half of the paper is devoted to the discussion about the influence of a large-scale balanced corpus on the linguistic study of the Japanese language, encompassing both traditional (corpus-based) and innovative (corpus-driven) research themes. Analyses of synonyms and collocations were presented as the examples of traditional research topics. Also, investigation about the inter-personal and/or situational differences of grammaticality judgment was presented as an example of innovative research topics. Lastly, issues about the corpus size were discussed from a point of view of the coverage by corpus of the total reading experience of a particular person. It turned out that the size of corpus should be considerably larger than one billion words if we want to construct a theory about the inter-personal difference of grammaticality judgments., application/pdf},
 pages = {13--28},
 title = {コーパス日本語学の可能性 : 大規模均衡コーパスがもたらすもの},
 volume = {22},
 year = {2007},
 yomi = {マエカワ, キクオ}
}