is an archaic kanazukai used to write Japanese during the Nara period. Its primary feature is to distinguish between two groups of syllables as discussed below that later merged together.

Syllables

Following are the syllabic distinctions made in Old Japanese.

a	i		u	e		o
ka	ki1	ki2	ku	ke1	ke2	ko1	ko2
ga	gi1	gi2	gu	ge1	ge2	go1	go2
sa	si		su	se		so1	so2
za	zi		zu	ze		zo1	zo2
ta	ti		tu	te		to1	to2
da	di		du	de		do1	do2
na	ni		nu	ne		no1	no2
ha	hi1	hi2	hu	he1	he2	ho
ba	bi1	bi2	bu	be1	be2	bo
ma	mi1	mi2	mu	me1	me2	mo1	mo2
ya			yu	ye		yo1	yo2
ra	ri		ru	re		ro1	ro2
wa	wi			we		wo

Those syllables marked in gray are known as Jōdai Tokushu Kanazukai.

Transcription

The two groups merged together by the 9th century. It predates the development of kana, and the phonetic difference is unclear. Therefore an ad-hoc transcription system is employed.

Syllables written with subscript 1 are known as type and those with subscript 2 as type .

There are several competing transcription systems. One popular system places a dieresis above the vowel: ï, ë, ö. This typically represents i₂, e₂, and o₂, and assumes that unmarked i, e, and o are i₁, e₁, and o₁. It does not necessarily have anything to do with pronunciation. There are several problems with this system.

• It implies a particular pronunciation, indirectly on the vowel.
• It neglects to distinguish between words where the 1 / 2 distinction is not clear, such as the /to/ in /toru/ as well as in /kaditori/.

Another system uses superscripts instead of subscripts.

Meaning

There are many hypotheses to explain the distinction. However, it is not clear whether the distinction applied to the consonant, vowel, or something else. There is no general academic agreement. See Old Japanese for more information.

Usage

A word is consistently, without exception, written with syllables from a specific group. For example, /kami₁/ "above" and /kami₂/ "god". While both words consists of an /m/ and an /i/, mi₁ can not substitute for mi₂ or vice versa. This strict distinction exists for all of the syllables marked in gray.

This usage is also found in the verb morphology. The conjugation is as follows:

Verb Class	Irrealis 未然形	Adverbial 連用形	Conclusive 終止形	Attributive 連体形	Realis 已然形	Imperative 命令形
Quadrigrade (四段)	-a	-i1	-u	-u	-e2	-e1

The verb /sak-/ "bloom" has conjugation class. Thus, its conjugation is as follows:

Verb Class	Irrealis 未然形	Adverbial 連用形	Conclusive 終止形	Attributive 連体形	Realis 已然形	Imperative 命令形
Verb	saka	saki1	saku	saku	sake2	sake1

It should be noted that before the Jōdai Tokushu Kanazukai discovery, it was thought that and shared the same form: -e. However, after the discovery, it became clear that was -e₂ while was -e₁.

Also, Jōdai Tokushu Kanazukai has a profound effect on etymology. It was once thought that /kami/ "above" and /kami/ "god" shared the same etymology, a god being an entity high above. However, after the discovery, it is known that "above" is /kami₁/ while "god" is /kami₂/. Thus, they are distinct words.

Man'yōgana chart

Following is a chart listing syllable and man'yōgana correspondences.

/a/	阿	安	婀	鞅	英	吾	足	嗚呼
/i/	伊	夷	怡	以	異	已	移	易	揖	因	印	壹	射	胆	五十	馬聲
/u/	汗	有	宇	于	羽	烏	紆	萬	雲	鬱	菟	鵜	卯	鸕	得
/e/	衣	愛	依	亞	哀	埃	榎	荏	得	可愛
/o/	意	於	隱	飫	淤	憶	應	邑	磤	乙
/ka/	加	可	賀	珂	迦	訶	箇	嘉	架	伽	歌	舸	軻	柯	介	甘	敢	甲	漢	干	葛	香	各	覺	髮	鹿	蚊	芳	歟
/ga/	奇	宜	何	我	賀	河	蛾	餓	峨	俄	鵝
/ki1/	支	吉	岐	伎	棄	枳	企	耆	祇	祁	寸	杵	來
/ki2/	歸	貴	紀	幾	奇	騎	綺	寄	記	基	機	己	既	氣	城	木	樹	黄
/gi1/	藝	祇	岐	伎	儀	蟻	𡺸
/gi2/	疑	宜	義	擬
/ku/	久	玖	口	句	群	苦	丘	九	鳩	倶	區	勾	矩	*	衢	窶	屨	君	訓	菊	來
/gu/	具	遇	求	隅	虞	愚	群
/ke1/	祁	家	計	鶏	介	奚	谿	價	係	結	稽	啓	兼	監	險	異
/ke2/	居	氣	希	擧	既	該	戒	階	開	愷	凱	慨	概	毛	食	飼	消	笥
/ge1/	下	牙	雅	夏	霓
/ge2/	義	宜	㝵	礙	皚
/ko1/	古	故	高	庫	祜	姑	狐	枯	固	顧	子	兒	籠	小	粉
/ko2/	己	許	巨	居	去	虚	忌	擧	*	拠	渠	金	今	近	乞	興	木
/go1/	胡	呉	候	後	虞	吾	誤	悟	娯

Development

The distinction between /mo₁/ and /mo₂/ is only made in the oldest text: Kojiki. After that, they merged into /mo/.

In later texts, confusion between type A and B can be seen. Nearly all of the A/B distinctions had vanished by the Classical Japanese period. As seen in early Heian Period texts such as Kogo Shūi, the final syllables to be distinguished were /ko₁, go₁/ and /ko₂, go₂/. After the merger, CV₁ and CV₂ became CV.

Bibliography

*