日本語の「音」を学んでいます【08】

日本語の「音」を学んでいます
Learning the “sounds” of Japanese Language
學習著日語的「音」
【08】 連濁・Sequential Voicing・連濁


☆

　內容 CONTENTS

1）連濁（れんだく）

　a) 濁點・Dakuten

　b) 濁音・Voiced sounds

　c) 連濁現象・Sequential voicing phenomenon

2）預測連濁・Predicting Sequential Voicing

　a) Lyman 定律・Lyman’s Law

　b) 右枝條件・Right Branch Condition

　c) 多過兩個成分的複合詞・Compound words with more than two components
　d) 局限・Limitations

3）あ゛

4）參考・References

☆☆.｡.:*･ﾟ*:.｡.☆☆.｡.:*･ﾟ*:.｡.☆☆.｡.:*･ﾟ*:.｡.☆☆.｡.:*･ﾟ*:.｡.☆☆

1） 連濁（れんだく）

　a) 濁點・Dakuten

出現在假名右上角的點點「 ﾞ 」叫做「濁點」。
The dots「 ﾞ 」that appear to the top right-hand corner of a kana is called “dakuten”（濁点）.

那是用來表示假名是濁化的附加符號。

It is a diacritic mark used to indicate that the kana is pronounced voiced.

濁聲：發聲時，聲帶振動的音（有聲）。
Voicing: When articulating, the vocal cords vibrates (voiced).

　b) 濁音・Voiced sounds

日語裡有四組濁音（だくおん）。
There are four groups of voiced sounds (濁音 (だくおん)) in Japanese.

這些濁音都是在對應的四組清音的右上角加上濁點形成的。
These voiced sounds are derived by adding dakuten to the top-right corner of the respective voiceless sounds.

和之前用的IPA標示的一些不同：
Difference in use of IPA notation from previous post: 

新的標記法 New notation	之前的標記法 Previous Notation
/s̆/	清齦顎擦音 /ɕ/ Voiceless alveo-palatal fricative /ɕ/
/j̆/	濁齒齦塞擦音 /dz/ Voiced alveolar affricate /dz/
/ts/	清齒齦塞擦音 /ts/ Voiceless alveolar affricate /ts/

（/j̆/ 在詞的開頭時，通常發 [dz]，不是詞的開頭時，則是 [z̆]。）

(For /j̆/, when it is in the word initial position, it tends to be [dz], when it is not in the word initial position, it tends to be [z̆]).

「じ」和「ぢ」，「ず」和「づ」是兩對同音不同形的假名。
「じ」and「ぢ」, 「ず」and「づ」are two pairs of syllabary with the same sound but different form.

（用羅馬字鍵盤打的話，輔音的部分是不一樣的。「じ」要打 “ji”，「ぢ」則是 “di”。「ず」要打 “zu”，「づ」則是 “du”。）

(If you type using romaji keyboard, the consonant is different. You type “ji” for「じ」, “di” for「ぢ」, “zu” for「ず」and “du” for「づ」.)

「血」讀作「ち」。

The word for blood（血）is read as「ち」.

在「鼻血」這個詞裡發生了連濁，讀作 [hanaj̆i]。

In the word “nosebleed”（鼻血）, rendaku occurs and it’s read as [hanaj̆i].

將 [hanaj̆i] 寫成平假名的時候，在「ち」（「血」的意思）上加上濁點，寫作「ぢ」。

When writing [hanaj̆i] as hiragana, you add the dakuten to「ち」(“blood”), writing it as「ぢ」.

雖然讀音一樣，但要根據濁化前的假名來寫對應的濁音「ぢ」，不能寫作「じ」。

Even though they have the same pronunciation, you have to write the voiced version of the syllabary「ぢ」, not「じ」.

「じ」和「ぢ」，「ず」和「づ」以前是對立的（發音不同），但現在在標準東京腔已經不對立了。

「じ」and「ぢ」,「ず」and「づ」 used to be contrasted (had different sound), but now they don’t in the Standard Tokyo accent.

じ・ず	ぢ・づ
	入れ知恵（いれぢえ）
	平調子（ひらぢょうし）
	三日月（みかづき）
一つずつ	つづく

　c) 連濁現象・Sequential voicing phenomenon

	A + B = AB”
か	平 　ひら	仮名 　かな	平仮名 　ひらがな
き	通し 　とおし	切符 　きっぷ	通し切符 　とおしぎっぷ
く	飛行機 　ひこうき	雲 　くも	飛行機雲 　ひこうきぐも
け	夫婦 　ふうふ	喧嘩 　けんか	夫婦喧嘩 　ふうふげんか
こ	迷 　まい	子 　こ	迷子 　まいご
さ	甘 　あま	酒 　さけ	甘酒 　あまざけ
し	小 　こ	島 　しま	小島 　こじま
す	花 　はな	寿司 　すし	花寿司 　はなずし
せ	蝉 　せみ	蝉 　せみ	蝉々 　せみぜみ
そ	星 　ほし	空 　そら	星空 　ほしぞら
た	剣 　けん	玉 　たま	剣玉・けん玉 　けんだま
ち	鼻 　はな	血 　ち	鼻血 　はなぢ
つ	紙 　かみ	作り 　つくり	紙作り 　かみづくり
て	手 　て	袋 　ふくろ	手袋 　てぶくろ
と	時 　とき	時 　とき	時々 　ときどき
は	ごみ	箱 　はこ	ごみ箱 　ごみばこ
ひ	人 　ひと	人 　ひと	人々 　ひとびと
ふ	露天 　ろてん	風呂 　ふろ	露天風呂 　ろてんぶろ
へ	両 　りょう	兵衛 　へえ	両兵衛 　りょうべえ
ほ	キラキラ	星 　ほし	キラキラ星 　きらきらぼし

發生連濁的都是組成複合詞的第二個成分。

The component that undergoes sequential voicing is the second member.

詞開頭的假名原本是四組清音之中的清音，組成複合詞並發生連濁後，變成了濁音（加了濁點）。

The first kana of the word is a voiceless sound from the four groups of voiceless sounds, after it forms a compound and undergoes sequential voicing, it becomes a voiced sound (dakuten is added).

有一個大津由紀雄在1980年時提出的條件，即發生連濁的部分必須是日本原有的詞（和語）。

There is a condition proposed by Otsu Yukio in 1980, that is the component that undergoes sequential voicing has to be a native Japanese word.

日語詞彙分類 Types of Japanese vocabulary	意思 Meaning	普遍的辨認方式（名詞） General way of recognising (nouns)
大和言葉 (やまとことば)	和語 Native Japanese Words	漢字訓讀 Kun-reading of Kanji
漢語 （かんご）	漢語（古漢語詞彙） Sino-Japanese (Classical Chinese words)	漢字音讀 On-reading of Kanji
外来語 (がいらいご)	外來語（非和語，非古漢語詞彙） Foreign words (non-Japanese word, non-classical Chinese)	用片假名書寫 Written with Katakana

這就是為什麼我會把上一篇專注於漢字的音讀和訓讀。
That’s why I dedicated the previous post to On-reading and Kun-reading of Kanji.

但「音讀」和「訓讀」也是一個要小心使用的標準，因為有各種例外。
But “On-reading” and “Kun-reading” is a criteria you have to use carefully, because there are all kinds of exceptions.

早期的借詞，例如葡萄牙語的「かるた」（注意它是用平假名寫的）可以發生連濁，即使它是外來語。

Early loan words, like Portuguese word「かるた」(note how it’s written in hiragana) can undergo sequential voicing despite it being a foreign word.

「漢字 (かんじ)」是漢語，「小籠包 (ショーロンポー)」是外來語，儘管兩個都能用漢字來寫。
「漢字 (かんじ)」is a Sino-Japanese word,「小籠包 (ショーロンポー)」is a foreign word, even though both can be written in Kanji characters.

「麻婆豆腐 (マーボーどうふ)」裡的「とうふ」會發生連濁，這可能跟「豆腐」被當作和語有關。
「とうふ」in「麻婆豆腐 (マーボーどうふ)」undergoes sequential voicing, it might have to do with how「豆腐」is seen as a Japanese word.

☆

2）預測連濁・Predicting Sequential Voicing

我想這部分會變得很理論性...
I guess this part will become very theoretical...



　a) Lyman定律・Lyman’s Law

Lyman定律主張當複合詞的第二個成員有濁阻礙音（塞音、擦音、塞擦音）時，連濁不會發生。

Lyman’s Law claims that sequential voicing does not occur when the second member of the compound has a voiced obstruent (stop, fricative, affricate).

　　日語的濁阻礙音：b, d, ɡ, z, z̆, d^z, j̆
　　Voiced Obstruents in Japanese: b, d, ɡ, z, z̆, d^z, j̆

　b) 右枝條件・Right Branch Condition

右枝條件主張連濁只會在能發生連濁的成分處在右枝時才會發生。
The Right Branch Condition (RBC) claims that sequential voicing only applies when a potential Rendaku segment is in a right branch constituent.





筆記・NOTES：

　　樹狀圖從最底層畫上去。

　　Draw the tree diagram from bottom to up.

　　需要考慮複合詞的意思。
　　You need to take into account the meaning of the compound.

　c) 多過兩個成分的複合詞・Compound words with more than two components

我會用辻村成津子的 “An introduction to Japanese Linguistics” 裡第三章音韻學的練習第十題的例子說明這個部分。

I will use Exercise 10 of Chapter 3 Phonology from Tsujimura Natsuko’s “An introduction to Japanese Linguistics” to explain this part.

C小題讓你在意思為「做一個架子放折紙用紙」的情況下，預測 “ori + kami + tana + tsukuri” 的連濁現象。
Part c is asking you to predict the sequential voicing for “ori + kami + tana + tsukuri”, given that the meaning is “making a shelf for paper for folding”.

おり　かみ　たな　つくり

    ori    kami    tana   tsɯkɯri

步驟・STEPS：

（1）
“ori” 和 “kami” 是一個複合詞，意思為「折紙 (成品)」或「折紙用紙」。
“ori” and “kami” forms a compound, having the meaning “origami (product)” or “paper for folding”.

　　＊記住，只有右枝的成分可以發生連濁。
　　＊Remember, only the right-branch constituent can undergo sequential voicing.

　　＊[kami] 裡沒有濁阻礙音，所以連濁可以發生（Lyman 定律）。
　　＊There is no voiced obstruent in [kami], hence sequential voicing can proceed (Lyman’s Law).

[oɾi] + [kami] —> [oɾiɡami]

 （2）
既然是放「折紙」的架子，那 [oriɡami] 會和 [tana] 結合成複合詞。
Since it is a shelf for the “paper for folding”, [oriɡami] will form a compound with [tana].

　　＊右枝條件和 Lyman 定律都要應用。

　　＊Both the RBC and Lyman’s Law applies.

[oɾiɡami] + [tana] —> [oɾiɡamidana]

（3）
最後 [oriɡamidana] 和 [tsɯkɯɾi] 結合成複合詞，意思為「做一個放折紙用紙的架子」。

Lastly,  [oriɡamidana] and [tsɯkɯɾi] form a compound, with the meaning “making a shelf for paper for folding”.

　　＊右枝條件和 Lyman 定律都要應用。

　　＊Both the RBC and Lyman’s Law applies.

[oɾiɡamidana] + [t^sɯkɯɾi] —>[oɾiɡamidanazɯkɯɾi]

根據題目給的語義，會畫出右下邊的樹狀圖。
According to the semantic meaning given by the question, you will get the tree diagram on the bottom right.

如果你考慮的意思是「用折紙做一個架子」的話，就會拿到左邊的樹狀圖和複合詞。

If you were to consider the meaning of “making a shelf with paper for folding”, you will get the tree diagram and compound on the left instead.


步驟・STEPS：

（1）[oɾi] + [kami] —> [oɾiɡami]

（2）

你「做一個架子」，所以 [tana] 和 [tsɯkɯɾi] 組合成複合詞。
You “make a shelf”, hence [tana] and [tsɯkɯɾi] form a compound.

　　＊右枝條件和 Lyman 定律都要應用。

　　＊Both the RBC and Lyman’s Law applies.

[tana] + [t^sɯkɯɾi] —> [tanazɯkɯɾi]

（3）
你「用折紙」「做一個架子」，所以 [oriɡami] 和  [tanazɯkɯɾi] 組合成複合詞。

You “use paper for folding” “to make a shelf”, hence  [oriɡami] and  [tanazɯkɯɾi] form a compound.

　　＊右枝條件和 Lyman 定律都要應用。

　　＊Both the RBC and Lyman’s Law applies.

　　＊  [tanazɯkɯɾi] 有濁阻礙音，所以根據 Lyman 定律，連濁不會發生。

　　＊There is a voiced obstruent in [tanazɯkɯɾi], hence according to Lyman’s Law, sequential voicing does not occur.

[oɾiɡami] + [tanazɯkɯɾi] —> [oɾiɡamitanazɯkɯɾi]

　d) 局限・Limitations

好的，為什麼我會用「主張」來表示 Lyman 定律和右枝條件？
Okay, why do I use “claim” for Lyman’s Law and RBC?

Lyman 定律不是沒有例外的。
Lyman’s Law is not without exceptions.

「梯子 (はしご)」在「縄梯子 (なわばしご)」裡能發生連濁，即使「梯子 (はしご)」裡有阻礙音。
「梯子 (はしご)」can undergo sequential voicing in「縄梯子 (なわばしご)」 even though there is a voiced obstruent in「梯子 (はしご)」.



右枝條件無法解釋日本人的姓氏裡的連濁現象發生或不發生的理由或規律。
RBC cannot explain the reason or rule in whether sequential voicing will or will not happen in family names of Japanese.

例如，「崎」在「山崎 (やまざき)」裡會發生連濁，但在「浜崎 (はまさき)」裡不發生連濁。
For example, 「崎」in「山崎 (やまざき)」undergoes sequential voicing but it does not undergo sequential voicing in「浜崎 (はまさき)」.

既然「規律」無法說通，那只能記住這些名詞怎麼念。
Since the “rules” fail you, the only thing you can do is to memorise how to read these names.



當規律預測說連濁可以發生，不代表連濁一定會發生。

When the rules for predicting whether sequential voicing can occur, it does not mean sequential voicing will always occur.

關於即使能發生連濁，卻不發生連濁的詞，可以參考這裏：
About words that do not undergo sequential voicing despite it being possible, you can read it up here:

RENDAKU AKA "SEQUENTIAL VOICING"ひ-ひ-ひ-ひとびと

https://www.tofugu.com/japanese/rendaku/

☆

3）あ゛

你可能在漫畫或者日本綜藝節目裡看過（或沒看過）有濁點的「あ」。
You might have (or have not seen) 「あ」with dakuten appearing in manga or subtitles in Japanese variety shows.

對山田涼介和「あ゛」的連接印象很深刻，所以把截圖翻了出來。
I have a strong impression of the link between Ryosuke Yamada and「あ゛」, so I rummaged for this screenshot.

（來源・Source: 嵐にしやがれ 2014.07.05）

就如同我在之前的文裡所提到的，元音和鼻音基本上都是濁（有聲）的，「あ゛」（還有其他有濁點的元音和鼻音）原則上應該不存在。
As I have mentioned in an earlier post, vowels and nasal sounds are all basically voiced and「あ゛」 (and any other vowels or nasal sounds with dakuten) technically should not exist.

但是我在字幕裡卻看到了這個現象，在漫畫裡也有這種文字的特殊使用現象。
But I saw this phenomenon in subtitles, there is also this unconventional usage of the writing in Manga.

有濁點的元音是用來表示聲帶強烈的振動，是把粗糙、沙啞的感覺，或者是人們感嘆是發出的聲音「視覺化」的一種做法。

Vowels with dakuten is used to indicate stronger vibration of the vocal cords, in attempt to “visualise” the hoarse, husky, rusty sound or the sound people make when they exclaim.

（山田君的情況則是辣到怪叫。）

(In the case of Yamada-kun, making random noises from the unbearable spiciness of the food).

「あ゛」不屬於正規書寫系統，而是一個嘗試把語言換成文字的使用現象。
「あ゛」do not belong to the proper writing system, but rather it is a phenomena where people tries to change language into writing.

文字是靈活的。
Writing is flexible.

日語的書寫系統用三種（不包括羅馬字）不同的正式的文字，而每一個系統都有自己的功能和常規的用法。
Japanese writing system uses three different (excluding Romaji) formal scripts, and each has its function and conventional uses.

然而，日語的書寫系統是很有彈性的，例如你能抱著某種目的，用平假名或片假名寫有漢字的詞。
Yet, the Japanese writing system is really flexible, like you can write words that have Kanji in Hiragana or Katakana, with certain intentions in mind.

好像被說服日語有必要保留至少三種書寫系統（漢字、平假名、片假名）了。
I feel I was convinced that Japanese needs to keep at least three writing systems (Kanji, Hiragana and Katakana).


☆

4）參考・References

Tsujimura, N. (2007). Phonology. In An introduction to Japanese Linguistics (2nd ed., pp. 50-58, 110). Blackwell.

weblio 辞書
http://www.weblio.jp

RENDAKU AKA "SEQUENTIAL VOICING"ひ-ひ-ひ-ひとびと

https://www.tofugu.com/japanese/rendaku/

Yahoo!知恵袋 — 「ず」と「づ」の使い方の違いを教えて下さい。
http://detail.chiebukuro.yahoo.co.jp/qa/question_detail/q1112812681

☆☆.｡.:*･ﾟ*:.｡.☆☆.｡.:*･ﾟ*:.｡.☆☆.｡.:*･ﾟ*:.｡.☆☆.｡.:*･ﾟ*:.｡.☆☆

語言和嘗試解釋語言的「理論」、「規律」等不是沒有例外的。

Language and the “theories”, “rules” etc. attempting to explain language are not without exceptions.

理論，說到底不是完美的。

Theories, after all are not perfect.

我覺得規律是用來幫助你理解，而不是用來束縛你的。

I think rules are supposed to help you understand, not restrict you.

☆

覺得這是最難寫的一篇。

I feel this is the hardest topic I have written on.

連濁其實是我最不願意寫，但又想探索的問題。
Rendaku is a question I don’t wish to write on, yet would like to explore.

說不願意寫是因為知道連濁很多很難解釋的例外。
Saying that I don’t wish to write on it is because I know there are a lot of exceptions that are hard to explain.

說想探索又是因為想看看裡面有什麼規律，以致會被稱作「現象」。
Saying that I would like to explore what are the patterns in it, so much so that it can be called a “phenomenon”.

我找到的例子都是「名詞＋名詞」的複合詞。
All the examples I found were those of “Noun + Noun” compounds.

針對連濁是否會發生在其他類型的複合詞（像是形容詞＋名詞那種，這也還是名詞複合詞）我還是沒有答案。

I still don’t have the answer to whether Rendaku can happen to other type of compounds (like adjective + noun, it still forms a nominal compound).

☆

語音和音韻的東西也差不多寫完了。
Stuff on phonetics and phonology are almost done.

最開始也不懂自己會寫出什麼，關於「音」。
At first, I did not know what I will end up writing, about “sounds”.

現在我有能力澄清了。
Now I have the ability to clarify it.

「語音學」說的是「語音」是怎麼產生的。

“Phonetics” talks about how “speech sounds” are produced.

「音韻學」說的是「語音的規律」。
“Phonology” talks about “sound patterns in speech sounds”.

從「音拍」開始，我都是在寫「音韻」的東西，而「語音」是寫「音韻」必要的知識和工具。
Starting from “mora”, I have been writing on “Phonology” and “Phonetics” is a fundamental knowledge and tool to write on “Phonology”.



這個學期結束後，可能會對之前寫的東西進行一些補充。
After this semester ends, I might add on to what I wrote previously.

我覺得這是一個很好的經驗。
I think this has been a very good experience.

嘗試理解，然後化成有條理的文。

Trying to understand, and turn it into a post with coherence.

（雖然我也不知道自己是否真的有邏輯。）
(Although I don’t know if I really have the logic.) 

日語音韻和日語詞法有關，也有一定的規律。
Japanese Phonology and Japanese Morphology is related, and there are patterns to it.

這樣的規律性是只存在於經過規範化的語言裡嗎？
Does this kind of pattern only exist in language that have been standardised?

☆

鏈接・LINKS

✿【00】出発点（しゅっぱつてん）

✿【01】「母音」と「長音」

✿【02】「子音」と「促音」

✿【03】「母音」、「子音」、「鼻音」

✿【04】拍

✿【05】無声化

✿【06】アクセント
✿【07】「音読み」と「訓読み」

☆

（完）