音声の相対音感 ~音声と音楽の同質性に関する一考察~ Relative Sense of Speech - A Study of Equality between Speech and Music -

この論文にアクセスする

この論文をさがす

著者

抄録

音声コミュニケーションには,話者・環境・聴取者に起因する音響歪みが不可避的に混入する。これら静的な非言語的歪みを数学的にモデル化し,そのモデルの上で,音響歪みを表現する次元を完全に失った音声の物理表象を提案している[1]。個々の音声事象の絶対的な物理特性は一切捨象し,音声事象間の関係のみを,全ての二事象間差異(コントラスト)の集合,即ち,ある幾何学構造として抽出する。この新しい物理表象は,構造音韻論の物理的実装として解釈されている。事象間のコントラストのみを捉える処理は,音楽の相対音感に類似した処理と考えられるが,本稿ではその提案表象を,言語学,心理学,言語障害学,神経生理学,脳科学,及び音楽学の観点から再度考察,解釈する。その中で,音素を音響空間内で定位する従来の方法論の是非について検討する。In speech communication, acoustic distortions are inevitably involved by speakers, channels, and listeners. In our previous study, these distortions were mathematically modeled, and on that model, a novel speech representation was proposed where the distortions cannot be observed [1]. Absolute properties of speech events are completely discarded and only their interrelations are extracted as a full set of phonic differences or contrasts. The set is mathematically equal to a certain geometrical structure. This new representation is considered as physical implementation of structural phonology. Extraction of contrasts between two events is viewed as a process similar to hearing music, i.e., relative pitch. In this paper, the new representation is reconsidered from viewpoints of linguistics, psychology, language disabilities, neurophysiology, brain science, and musicology, Here, the conventional paradigm where a phoneme is localized absolutely at a certain point in an acoustic space is also reconsidered.

In speech communication, acoustic distortions are inevitably involved by speakers, channels, and listeners. In our previous study, these distortions were mathematically modeled, and on that model, a novel speech representation was proposed where the distortions cannot be observed [1]. Absolute properties of speech events are completely discarded and only their interrelations are extracted as a full set of phonic differences or contrasts. The set is mathematically equal to a certain geometrical structure. This new representation is considered as physical implementation of structural phonology. Extraction of contrasts between two events is viewed as a process similar to hearing music, i.e, relative pitch. In this paper, the new representation is reconsidered from viewpoints of linguistics, psychology, language disabilities, neurophysiology, brain science, and musicology, Here, the conventional paradigm where a phoneme is localized absolutely at a certain point in an acoustic space is also reconsidered.

収録刊行物

  • 情報処理学会研究報告音声言語情報処理(SLP)

    情報処理学会研究報告音声言語情報処理(SLP) 2005(127(2005-SLP-059)), 211-216, 2005-12-22

    一般社団法人情報処理学会

参考文献:  40件中 1-40件 を表示

各種コード

  • NII論文ID(NAID)
    110003494758
  • NII書誌ID(NCID)
    AN10442647
  • 本文言語コード
    JPN
  • 資料種別
    Technical Report
  • ISSN
    09196072
  • NDL 記事登録ID
    7768316
  • NDL 雑誌分類
    ZM13(科学技術--科学技術一般--データ処理・計算機)
  • NDL 請求記号
    Z14-1121
  • データ提供元
    CJP書誌  NDL  NII-ELS  IPSJ 
ページトップへ