Answer Home & Living 5 min read

How many kana there are depends on which list is counted

Japanese is written partly in hiragana and katakana, and a learner's chart shows 46 kana, but the official kana list in force since 1986 runs to 108 entries by our count. Unicode encodes far more, including 285 historic hiragana variants.

Eva Chin
Business & Chinese Culture Correspondent
Published 19 Oct 2026, 2:18 PM (SGT)
Share:
Hiragana characters brushed in black ink on white paper Hiragana characters brushed in black ink on white paper Photo by Mikhail Nilov on Pexels
Advertisement

Japanese is written partly in two kana scripts, hiragana and katakana, in which each sound takes one kana or a pair of them. How many kana there are depends on what is counted. By our count, the rule in force today, the 1986 cabinet notice on modern kana usage, lists 108 entries, 36 of them two-kana combinations. A learner's chart shows 46: the 45 kana that modern spelling places in the fifty-cell grid, plus ん. Charts that keep ゐ and ゑ, which left everyday writing in 1946, show 48. Fifty is the number of cells.

The Japan Foundation's 2024 survey of Japanese-language institutions abroad ranks Indonesia second, with 732,914 learners at responding institutions, and puts Thailand, Vietnam and Myanmar in its top ten too; the ranking covers countries and regions, Taiwan among the latter. Romanisations below are modified Hepburn.

Everyday kana shed ゐ and ゑ under a 1946 cabinet notice

Japan's National Language Council recommended new kana rules on 21 September 1946, and the cabinet issued them as a notice on 16 November, according to the Agency for Cultural Affairs' record of the council. The council's text says that ゐ, ゑ and を are written い, え and お, except for the particle を. Kanji forms are a separate story, told in our guide to old and new kanji forms.

The directive issued with 昭和61年内閣告示第1号 on 1 July 1986 abolishes the 1946 notice, and the 1986 list again leaves out ゐ and ゑ. The particle keeps を, and the notice's examples include やむをえない and てにをは. ぢ and づ also survive, in words such as ちぢむ and はなぢ.

Present-day kana fill all but five cells of the gojūon grid

The gojūon, or fifty-sound table, sets ten rows against the five vowels a, i, u, e and o. Modern kana fill 45 cells. The Y row holds only や, ゆ and よ, and the W row only わ and を, which leaves yi, ye, wu, wi and we empty. ん sits outside the grid. The 1986 notice's own table prints よ and を in the o column.

Until 19 October 2026 our kana chart packed the Y and W rows to the left, so よ sat under u and を under i. Its transcriber wrote kawaī, nīgata and omō, and it lacked ヴュ, so ヴュー came out as "vuュ". Its Kunrei-shiki and Nihon-shiki paths failed on all 32 extended katakana it carried.

The revision committed that day puts each kana under its vowel. Tick "ゐ / ゑ" under Include and the W row reads わ, ゐ, a blank, ゑ, を. The transcriber now writes omou, kawaii and niigata by default, with macrons as an option, the Extended group carries 33 combinations, and a system with no spelling for a kana returns it in brackets, as in [ファ].

Voiced, small and combined forms bring the current list to 108 entries

The notice prints no total and names no row voiced or semi-voiced; the grouping and sum below are ours, from its first section.

GroupEntries
Plain kana in the grid45
Voiced (g, z, d, b rows)20
Semi-voiced (p row)5
Two-kana combinations (きゃ and the like)36
ん1
っ, the mark for a doubled consonant1
Total108

The notice underlines six of the 108 and reserves them for the special cases in its second section: ぢ, づ, を and the three combinations ぢゃ, ぢゅ and ぢょ. The learner's 46 is the first row plus ん; 48 adds ゐ and ゑ.

Advertisement

Loanword katakana adds 33 combinations from a separate cabinet notice

A cabinet notice of 28 June 1991 on writing loanwords lists combinations the 1986 list lacks, 13 in its first table and 20 in its second. The first holds ファ and ティ; the second holds ヴァ and ヴュ, the last for words like インタヴュー. Where second-table kana are not needed, the notice permits plainer ones, with ヴァ to バ as its own example.

In 2019 a statute, 平成31年法律第7号, amended the law naming Japan's overseas missions, and from 1 April that year it respelled Cabo Verde and Saint Kitts and Nevis without ヴ, as the two revisions on e-Gov show.

Romaji spellings of the same kana differ across three systems and the passport rule

Hepburn, Kunrei-shiki and Nihon-shiki remain in print. The cabinet notice of 22 December 2025 abolished the 1954 notice and uses Hepburn-type spellings in its main table, with ヲ as o. For a long i-row vowel it calls the doubled spelling the usual one, kawaii and niisan. In おもう the う is part of the verb ending, so it is spelled omou.

Passports follow their own rule, which the Kanagawa passport centre explains: a final long o is written, so 高遠 is TAKATOO. Sapporo's own site writes its beer ヱビス and spells it YEBISU. Our guide to Japanese romanisation and names covers the rest.

Hentaigana outnumber the standard hiragana letters in Unicode

Counted from the Unicode 18.0 data files, the main Hiragana block, U+3040 to U+309F, has 93 of its 96 code points assigned, and 86 of those are hiragana letters. The Katakana block, U+30A0 to U+30FF, has all 96 assigned, and 90 are katakana letters. The rest are marks, such as the voicing and iteration marks, plus one digraph in each block. Further hiragana letters, small ゐ among them, sit outside the main block.

Hentaigana, historic variant forms of hiragana, are a separate set. Unicode 10.0, released in June 2017, added "a set of 285 Hentaigana characters", and all 285 date from that version. They cover 48 syllable names, and KA has the most forms, 12. These figures do not add to the block counts.

Others tabulated this before us. John D. Cook mapped the 46 basic kana into the two 96-code-point blocks in a 2022 post, and Wikipedia's block articles list counts by version. Ours are a cross-check against the current files.

Katakana now has a code point for every cell of the grid

Katakana has had ヰ and ヱ since Unicode 1.1, and Unicode 14.0, released on 14 September 2021, added KATAKANA LETTER ARCHAIC YI, YE and WU at U+1B120 to U+1B122. Hiragana has ARCHAIC YE (U+1B001) and ARCHAIC WU (U+1B11F) but no ARCHAIC YI, so it reaches 49. Encoding a letter says nothing about whether the sound was ever distinct.

Unicode 18.0, announced on 16 September 2026, added seven more kana code points. Many phones may not yet display them.

Code pointUnicode name
U+1B123HIRAGANA DIGRAPH KOTO
U+1B124KATAKANA DIGRAPH TOKI
U+1B125KATAKANA DIGRAPH TOTE
U+1B126KATAKANA DIGRAPH YORI
U+1B127KATAKANA LETTER ALTERNATE NE
U+1B128KATAKANA LETTER ALTERNATE WI
U+1B168KATAKANA LETTER SMALL ARCHAIC YE

The new small archaic ye is katakana too. Hiragana's missing yi remains the one empty cell in either script.

Advertisement
Eva Chin
Business & Chinese Culture Correspondent

Eva Chin covers business and commerce in Southeast Asia for RECATOOLS, alongside Chinese cultural practice and education.

View author profile → · Editorial policy

Corrections policy

Advertisement