日本語ノンヘデッド複合語における音素上の偏り 「ワープロ」,「あれこれ」,「めちゃくちゃ」の共通点について

ローランス・ラブリュヌ (ボルドー第3大学)



宮腰 幸一 (筑波大学)


素性階層理論と音声素性 ―日本語の有声化とコーダ鼻音化についての試み―

黒田 成幸 (カリフォルニア大学サンディエゴ校/国際高等研究所)


Three Types of Imperatives: Japanese/English Imperatives and the Scale of Potential/Actual-Type

Hideki Mori
(Osaka University)

The purpose of this study is to consider imperatives in Japanese and English from the perspective of general linguistics. The idea that imperatives can have a negative implication has been repeatedly exploited to classify them. As far as English imperatives are concerned, this criterion may be necessary and sufficient. They can thereby be divided into two groups: imperatives that carry negative implications and those that do not. I will demonstrate, however, that this does not hold true for Japanese imperatives. As evidence for this, I will discuss the compatibility of the Japanese auxiliary verb miro with imperatives. It is claimed that imperatives have at least three basic types, which are distinguished by different degrees of compatibility with the verb in question: optional, obligatory and unacceptable. To give a principled account of this fact, I propose a scale of the propositional content of imperatives, which has two polar opposites: potential-type and actual-type. On this scale, imperatives are classified according to whether they rely on the speaker's direct experience. Not only does the proposed scale allow us to explain different degrees of compatibility of miro, but it can also offer several refinements to existing knowledge about imperatives. Since imperatives, whether affirmative or negative, are located somewhere on the scale, no assumption of the so-called idiosyncrasy of negative imperatives is needed in the present paper. The discrepancy in the scale between Japanese and English imperatives in turn makes their typological differences clear.

A Comparative Study of Haya, Ankole, and Tooro Tone Systems in Connection with Tone Loss in Tooro

Shigeki Kaji
(Kyoto University)

In western Uganda, some closely related Bantu languages such as Ankole, Kiga, Tooro and Nyoro are spoken. These languages are sometimes referred to as the Runyakitara language group. This group also includes Haya, a Tanzanian language which is spoken just to the south of Ankole. When we look at the tone systems of these languages, there are some striking differences. The most obvious difference is that, whereas Haya, and also Ankole to a certain extent, retain a relatively old tonal system (Kiga and Nyoro data lacking), in which the disyllabic -HL, -LH and -LL noun stems are differentiated, Tooro, which is spoken to the north of Ankole, has almost completely lost the original tonal system: the penultimate syllable of the word is always high-pitched in isolation. The author tries to explain how the Tooro system, which phonologically lacks tone, has come into being, by analyzing the differences which exist between the Haya system and the Ankole system. The Haya system is the oldest among these languages, and the Ankole system can be characterized as moving from the Haya system toward the Tooro system. We find in Ankole itself some signs of change, such as non-clear distinction of the underlying -HL and -LH patterns, and high tone loss in a number of words.

日本文の処理におけるスクランブル効果の諸問題 ―Koizumi and Tamaoka(2004)に対するMiyamoto and Nakamura(2005)のコメントへの回答

玉岡賀津雄 (広島大学)
小泉政利 (東北大学)

Koizumi and Tamaoka(2004,以下KT)の実験結果に対して,Miyamoto and Nakamura(2005,以下MN)が寄せたコメントについて,3つの視点から回答した.第1に,KTの実験条件ではMNが指摘しているような再解析が起こらないことを示唆する経験的証拠を提示し,少なくともKTの実験においては,統語構造の複雑さが文正誤判断課題の反応時間と関係するという仮定が成り立つことを再確認した。第2に,頻度について,語彙,統語,共起の3種類があることを説明した.語彙頻度についてはKT(2004)の実験では比較条件で一様であるため,影響がない.統語頻度については,KTが使用した実験刺激の頻度をMNが示している(MN, Table 2, p. 121)が,これが実験統制に用いられる通常の方法であるにもかかわらず,結果を不十分としているのは不適切な議論であることを指摘した.MNの文完成課題については,興味深い結果を得ているものの,オフラインの結果であるため,それを支持するためのオンラインの実験が必要であろう.共起頻度については,今後の研究を待つことになろう.第3に,文正誤判断課題と自己制御読みに関して,MN自身も指摘しているように,自己制御読みについては複雑な文でない限り有意なスクランブル効果が観察されていない.本稿では,近年行われた文正誤判断課題の実験がスクランブル効果を一貫して観察していることを示した.その上で,自己制御読み実験では,キー押しのための運動が文処理に影響すること,読みのために与えられる課題が「二重課題法」のように機能して,文処理以外に過度な記憶負荷をかけていることなどを指摘した.

A Numeral Denoting ‘One’ in Ainu

Hideo Kirikae
(Hokkai Gakuen University)

This paper attempts to show that in Ainu the number ‘one’ can be referred to by the proclitic ar which means ‘one of a pair’ (as-siki ← ar-siki ‘his eye’) or ‘half’ (ar-kewtemu ‘half of his heart’), and also functions as an intensifier (ar teynep ‘a true baby’). Although ar is qualified to be a numeral semantically, it has not been treated as such in traditional descriptions.

It is arguable that in general a numeral denoting ‘one’ can be derived from the concept of ‘a half of two’ and that a typical case of illustrating this process is observed in Ainu.