計量国語学
Online ISSN : 2433-0302
Print ISSN : 0453-4611
論文A
言語類型論への多重対応分析による アプローチ(1)
Tsunoda, Ueda, and Itoh (1995a) の統計科学的再考察
小野 洋平吉野 諒三林 文ホイットマン ジョン
著者情報
ジャーナル オープンアクセス

2017 年 31 巻 3 号 p. 189-204

詳細
抄録
This paper uses Multiple Correspondence Analysis (MCA) to re-analyze the linguistic data studied in Tsunoda, Ueda, and Itoh (1995a). Tsunoda et al. (1995a) applied cluster analyses to the word order parameters in Tsunoda, Ueda, and Itoh’s (1995b) database and identified as the main component the word order parameter category "prepositional," thus positing a basic dichotomy between prepositional languages versus adpositionless and postpositional languages. We reanalyze the same data using MCA, together with certain clustering techniques, applied to both the languages and the linguistic parameters in Tsunoda et al.’s (1995b) database simultaneously. This approach confirms the significance of word order parameters involving adposition but suggests a reconsideration of Tsunoda et al.’s (1995a) classification. We find a dichotomy between a cluster of "head-initial" (prepositional, verb-object, and noun-genitive) parameter categories versus a cluster of “head-final” (postpositional, object-verb, and genitive-noun) parameter categories similar to those often grouped together by linguists on the basis of informal cross-linguistic observations. We also find that Noun Phrase internal word order parameters, excepting parameter governing the order of noun and genitive, cluster separately from the others.
著者関連情報
© 2017 計量国語学会

この記事はクリエイティブ・コモンズ [表示 - 非営利 - 改変禁止 4.0 国際]ライセンスの下に提供されています。
https://creativecommons.org/licenses/by-nc-nd/4.0/deed.ja
次の記事
feedback
Top