Hiragana and katakana chart

Hiragana and katakana are Japan's two syllabaries, and they write exactly the same 46 sounds using two different sets of shapes. Learn hiragana first: it carries the grammar, the verb endings and the particles that hold a sentence together, while katakana handles foreign words, names and sound effects. Neither one is a puzzle.

They're two alphabets you memorize, and the memorizing is smaller than it looks, because the extra sounds reuse the same 46 shapes with marks added on top. Here's the full chart for both scripts, base kana, dakuten, yōon and the combinations Japanese invented for foreign sounds, followed by the look-alike pairs behind almost every beginner misread.

The hiragana chart

Hiragana is the foundation. Read it in rows: each row is one consonant crossed with the five vowels a, i, u, e, o. The traditional name for this grid is gojūon, "fifty sounds", though modern Japanese only keeps 46 of them.

The gaps are real, not typos. The y row only exists with a, u and o. The w row lost everything but わ and を, and two kana, ゐ (wi) and ゑ (we), were retired from everyday writing by the 1946 spelling reform, which is how "fifty sounds" ends up being 46 characters.

Four cells trip up people who assume the pattern holds all the way across. し is shi, never "si". ち is chi, never "ti". つ is tsu, never "tu". ふ sits between an f and an h, closer to blowing out a candle than to English "foo", and it gets romanized fu. And ん is the only kana that isn't a consonant plus a vowel: it stands alone, takes a full beat, and never starts a word.

Dakuten and handakuten

Two small strokes at the top right of a kana (dakuten, 濁点) turn its consonant voiced. か ka becomes が ga. A small circle instead (handakuten, 半濁点) turns the h row into p. No new shapes to learn, just 25 more sounds.

ぢ and づ sound identical to じ and ず in modern Japanese. You'll meet them almost only inside compounds where an original ち or つ picks up voicing: 鼻血 (はなぢ, hanaji, nosebleed), 続く (つづく, tsuzuku, to continue).

Yōon: the small ya, yu and yo

Take any kana from the i column, follow it with a shrunken ゃ, ゅ or ょ, and the two fuse into a single syllable. Size matters. きゃ is one beat, kya, while きや at full size is two beats, ki-ya, and that difference changes words.

Add it up: 46 base kana, 25 with marks and 33 combinations, so 104 written syllables out of 46 shapes. The same arithmetic runs again for katakana below.

The small tsu and long vowels

っ is a shrunken つ and it makes no sound of its own. It doubles the consonant that follows and inserts a beat of silence before it: きって (kitte, stamp) against きて (kite, come). Miss it and you say a different word.

Hiragana stretches a vowel by writing another vowel kana after it, never by using a stroke. おかあさん is okāsan, mother. とうきょう is Tōkyō, spelled out as to-u-kyo-u. That last one catches everyone, because the o sound is written with a う.

The katakana chart

Same 46 sounds, angular shapes, different job. Katakana writes loanwords, foreign names, scientific names, onomatopoeia and emphasis, which makes it the script you actually run into first in the wild: menus, packaging, shop signs, manga sound effects. Writing your own name in Japanese uses katakana too, and there's a whole method to that.

ヲ is a curiosity. The object particle is always written in hiragana, so katakana ヲ shows up mostly in stylized signage, old telegrams and the occasional game menu. Learn to recognize it, don't bother writing it.

The small ッ works exactly like っ: カップ is kappu, cup. ベッド is beddo, bed.

ー is a vowel, not a dash

Read this bit twice, because it's the most misread mark on the whole chart. ー is a character called chōonpu (長音符), and it holds the vowel in front of it for an extra beat. ラーメン is rāmen, four beats, not "ra men" with a hyphen stuck in the middle. コーヒー is kōhī. ケーキ is kēki, cake.

Drop it and the word changes. ビル is biru, a building. ビール is bīru, beer. One stroke separates ordering a beer from ordering an office block.

In vertical text the mark is drawn vertically, which is why beginners reading manga sometimes mistake it for part of the speech bubble or a panel border. It isn't. It's a letter.

Hiragana doesn't use ー in normal writing. It stretches vowels with vowel kana instead, the おかあさん way.

Sounds Japanese borrowed later

Modern Japanese needs syllables the old grid never had, so katakana builds them by pairing a kana with a small vowel. These combinations exist in katakana only.

ヴ is optional in practice. Plenty of publishers still write バイオリン for violin, so you'll see both spellings for the same word.

The pairs that cause misreads

Four small groups of katakana account for most decoding errors, and a single mnemonic handles the two worst. Look at where the strokes start. シ and ン sweep upward from the bottom left, the way their hiragana cousins し and ん are written. ツ and ソ come down from the top, the way つ and そ are written.

Hiragana has its own confusable set, milder but real: ね, れ and わ differ only in how the tail resolves; る closes into a loop where ろ stops short; さ has one crossbar and き has two; は and ほ differ by one horizontal line.

Three kana that lie about their sound

These aren't shape problems. They're reading problems, and no chart shows them.

1. は is read wa when it marks the topic. 私は学生です is watashi wa gakusei desu. Everywhere else it's ha, as in はな (hana, flower). This is also why こんにちは comes out as konnichiwa despite ending in は: it's a frozen particle from an old phrase.
2. へ is read e when it marks direction. 東京へ is Tōkyō e. On its own it's he, as in へや (heya, room).
3. を is written wo and pronounced o. It survives almost exclusively as the object particle: パンを食べる is pan o taberu, I eat bread.

Bonus trap while you're here: the u in です and ます goes nearly silent. です comes out closer to des, ます closer to mas. The kana are still written in full.

How to learn all of this, in order

Charts don't teach. Order and repetition do. Here's a sequence that front-loads the rows carrying the most common words instead of marching alphabetically to the end.

1. The five vowels on their own, first. あ い う え お. Every other kana is one of these with a consonant welded to the front, so getting them wrong poisons everything downstream.
2. Then the k, s, t and n rows. Those four plus the vowels come to 25 characters, and they carry an enormous share of everyday words. One row per sitting: say it out loud, cover the romaji, say it again.
3. Then h, m, y, r, w and ん. The tail of the chart is quicker than the head, because two of those rows are incomplete.
4. Write each kana by hand once or twice as you go. Not calligraphy. The hand remembers stroke order in a way the eye doesn't, and it's what stops し and つ from blurring together three weeks later.
5. Add dakuten and yōon at the very end. They cost nothing new. If you can read か and は you can already read が and ぱ, and きゃ is just き plus a shrunken ゃ.
6. Then run katakana from the top, and drill it harder than you drilled hiragana. Beginner material leans on hiragana, so katakana fades unless you go looking for it. Product packaging is the cure.

Budget 15 to 20 minutes a day and about a week per script to recognize every character without stalling. Reading at a comfortable pace, where you're taking in words instead of decoding letters, takes another month or two of actually reading things. The shapes come fast. The speed comes from mileage, and there's no version of this where you skip the mileage.

One rule worth being strict about: drop romaji as soon as you can sound out a full row unaided. Romaji and kana compete for the same word, and the eye always takes the easier path, which is exactly how people spend six months "learning hiragana" without ever reading a sentence without training wheels. Once the kana are in, the three scripts of written Japanese stop being a wall and start being a system.

Where the kana actually pay off

Start using them the day you finish the k row, on things you genuinely want to read.

Menus give you the fastest wins because so much of a Japanese menu is katakana. カレー, ラーメン, サラダ, ビール. The hiragana half is where the useful words hide: おすすめ (recommended), 大盛り with its おおもり reading, とんかつ. Point at what you can now half read and work out the rest of the menu from there. Counting to ten closes the gap on prices and portions.

Station signs and exits come next. 出口 and 入口 carry furigana on a lot of signage, and furigana is hiragana printed small above kanji, so the kana you just learned are literally the pronunciation key for characters you haven't studied yet.

Manga is the third one, and it's the reason a lot of people start. Covers, chapter titles and children's series print furigana over most kanji, sound effects are pure katakana, and dialogue is heavy on hiragana grammar. That's why reading raw manga turns into practice instead of frustration once the kana are solid.

Where a camera translator earns its place is the middle stretch, when you can pronounce a line but not understand it. That gap is real and it lasts months. Point Yomi at the page or the sign and the English lands over the Japanese in place, with the romaji alongside, so you can check the reading you just worked out rather than guess and move on. It needs a connection to do that, so it's a tool for the menu in front of you, not a replacement for the chart above. Treat it as the thing that keeps you reading real Japanese while the kana are still settling, and it does its job. Treat it as a substitute for learning the 46 shapes, and you'll be squinting at a phone in six months for words you should be able to read cold.