A quantitative analysis of the development of embedded clauses in a large corpus of children's writing

2023 Fiscal Year Final Research Report

• Project/Area Number: 19K23068
• Research Category: Grant-in-Aid for Research Activity Start-up
• Allocation Type: Multi-year Fund
• Review Section: 0102:Literature, linguistics, and related fields
• Research Institution: University of Tsukuba
• Principal Investigator: Imada Mizuho 筑波大学, 人文社会系, 助教 (10579056)
• Project Period (FY): 2019-08-30 – 2024-03-31
• Keywords: 児童作文 / コーパス言語学 / 統語的複雑性 / 節境界

Outline of Final Research Achievements

We developed a linguistic resource for the study of children's compositions, and used it to study syntactic complexity. First, we modified an existing composition corpus and further annotated clause boundary labels. We then analysed syntactic complexity and observed that while dependency distance is suppressed in natural language, rather long dependency relations are often observed in short sentences. This result suggests that the development of children's syntactic competence has two aspects: complexity and rationalisation. The analysis of clause types confirms that the change from coordinate to subordinate structures and from spoken to written language occurs with increasing school age.

Free Research Field

日本語学

Academic Significance and Societal Importance of the Research Achievements

データを基盤とした言語の実証的研究においてはデータと分析手法の両方が必要となるが、児童の統語能力の発達は、データの入手困難性と、語彙情報に比べて構文情報の分析の技術的要求の高さから、比較的活発に研究されてこなかった領域と言える。本研究の成果は、児童の統語的能力の発達という言語学的な課題に対して、データを用いた実証的な知見を提示する。また、この研究は言語学における言語資源の活用領域を語彙研究から構文研究への拡大を促進すること、および自然言語における統語構造の数学的特性についての理解を深めることに貢献する。