A Study of the Accent of Kyengsang Korean(Fostering Joint International Research)
| Project/Area Number |
15KK0041
|
|---|
| Research Category |
Fund for the Promotion of Joint International Research (Fostering Joint International Research)
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Linguistics
|
|---|
| Research Institution | Tokyo University of Foreign Studies |
|---|
Principal Investigator |
Ito Chiyuki 東京外国語大学, アジア・アフリカ言語文化研究所, 准教授 (20361735)
|
|---|
| Research Collaborator |
Albright Adam マサチューセッツ工科大学, 言語学科, 教授
Feldman Naomi マサチューセッツ工科大学, 言語学科, 客員准教授
Kenstowicz Michael マサチューセッツ工科大学, 言語学科, 教授
|
|---|
| Project Period (FY) |
2016 – 2018
|
| Project Status |
Completed (Fiscal Year 2018)
|
| Budget Amount *help |
¥12,220,000 (Direct Cost: ¥9,400,000、Indirect Cost: ¥2,820,000)
|
|---|
| Keywords | 韓国朝鮮語 / 慶尚道方言 / 延辺朝鮮語 / アクセント / 類推変化 / 繰り返し学習モデル / 歴史言語学 / 中期朝鮮語 / 音韻論 |
|---|
| Outline of Final Research Achievements |
In this study, I tried to explain the historical development of the distinctive pitch accent in two Korean dialects (Kyengsang, Yanbian) from Middle Korean by constructing computational models. Given that the accent patterns in the Korean dialects differ depending on word class (native/Sino-Korean/loanwords), I first constructed models to test to what extent these classes can be distinguished based solely on phonological factors. The constructed models successfully captured these lexical distinctions quite well.
Second, I constructed models of historical accent changes in disyllabic Sino-Korean words of Kyengsang/Yanbian by using an iterated learning procedure. I assumed that a certain proportion of Sino-Korean morphemes that compose Sino-Korean words lost their underlying accent and so speakers assigned them an accent based on phonological knowledge they induced from the data of the previous generation. The models reflected the historical development of the accent quite accurately.
|
| Academic Significance and Societal Importance of the Research Achievements |
韓国朝鮮語の語彙種（固有語・漢字語・借用語）が、音韻論的条件のみによって区別が可能であることを、音韻論ベースのモデルを構築することにより明らかにした研究はこれまでにない。また、繰り返し学習モデルを用いた先行研究では、主として人工的な言語が研究対象とされているが、本研究は同手法を、韓国朝鮮語諸方言におけるピッチアクセントの歴史的発展という、実際の歴史変化に対して応用した点において、新規性があると言える。
|
Report
(4 results)
Research Products
(7 results)
