Japanese Writing System: 3 Scripts, Which to Learn First

Japanese uses three scripts that share the work of one sentence, and none of them is an alphabet. Two are syllabaries, ひらがな (hiragana) and カタカナ (katakana), with 46 basic characters each.

The third is 漢字かんじ (kanji), the characters borrowed from Chinese. A syllabary is not an alphabet, because each character stands for a whole syllable. か (ka) is “ka,” and you never assemble it from a k and an a.

The three are not rival versions of the same thing, and you do not pick one. Every ordinary sentence uses all three at once.

Here is one ordinary sentence with all three scripts in it:

私は毎朝コーヒーを飲みますわたしはまいあさコーヒーをのみます (watashihamaiasakoohiiwonomimasu)

I drink coffee every morning.

Each word in the sentence takes the script that matches its job. A grammar ending stays in hiragana, so a change of script often shows where the next word starts. A verb's kana ending is the exception, because it belongs to the verb, not to a new word.

Kanji carry the meaning. Hiragana carry the grammar wrapped around that meaning. Katakana mark the words that arrived from somewhere else.

This field note takes one real sentence apart to show the split, and explains why three scripts make Japanese easier to read rather than harder. Then it gives you the order to learn them in: hiragana, then katakana, then kanji. When a script needs a week of its own, this note points you to the guide that covers it.

If you came looking for the Japanese alphabet, there is not one. What you want is a kana chart. Hiragana and katakana are syllabaries, each with 46 basic characters, and the two spell exactly the same set of sounds.

That last part surprises people. か (ka) and カ (ka) are both “ka.” ね (ne) and ネ (ne) are both “ne.”

They are not a simpler set and a harder set. It is one sound system written in two hands, kept apart so each can say something different about the word it spells.

Kanji work on a different principle. A kanji stands for a meaning first and a sound second, which is why one character can be read more than one way. 山やま (yama) is a mountain on its own and still a mountain inside a longer word, even when the reading changes.

Each script has its own home, and after a few weeks of reading you will spot them without thinking about it. Hiragana is the connective tissue: particles, verb and adjective endings, and native words that have no everyday kanji. It is what turns a pile of meanings into a sentence.

You meet it on the tail of every verb. Here it is doing the polite ending on “eat”:

食べますたべます (tabemasu)

I eat. (polite)

Katakana marks words that came in from outside: loanwords, foreign names, and most of the product words on a shelf. It has a second job that surprises you at first.

It also writes sound effects, plant and animal names in scientific writing, and words set in katakana purely for emphasis. English uses italics in roughly the same way.

Two you will read on day one are コーヒー (koohii) (coffee) and パン (pan) (bread), borrowed generations ago and still written in katakana today.

Kanji carry the heavy vocabulary: nouns, and the meaning-bearing stem of most verbs and adjectives. They're the words you'd look up in a dictionary. Everything hiragana adds around them — は (ha), を (wo), ました (mashita) — tells you what those meanings are doing to each other.

Here is a sentence a beginner meets in the first month, written the way any Japanese adult would write it:

私は毎朝コーヒーを飲みますわたしはまいあさコーヒーをのみます (watashihamaiasakoohiiwonomimasu)

I drink coffee every morning.

Six pieces, three scripts, and every switch is doing a job. Read the table one row at a time: each row is one piece of that sentence and the reason it wears the script it wears.

Particles like は (ha) and を (wo) are grammar rather than content, so they stay in hiragana. A verb splits in two, with the kanji holding the meaning and the hiragana carrying the polite ending.

Japanese is written without spaces between words. That sounds like a problem until you notice what the scripts are doing: each change of script marks a seam.

Take a second sentence and watch the seams appear:

今日はパンを買いましたきょうはパンをかいました (kyouhapanwokaimashita)

I bought bread today.

There are no gaps in that line, and you do not need any. 今日きょう (kyou) is kanji, は (ha) is hiragana, パン (pan) is katakana, and を (wo) is hiragana again. 買いましたかいました (kaimashita) starts in kanji and ends in hiragana.

Four changes of script give you four word boundaries for free.

So the honest answer to “why three?” is not that Japanese ended up with three alphabets doing one job badly. The division of labor does what spaces do in English.

Text written in kana alone, the way books for very small children are printed, gives you none of those seams. You have to find the word breaks yourself.

One line of history explains all three. Japanese was a spoken language with no writing of its own when Chinese characters arrived, and for centuries kanji did everything.

Writers then began using certain kanji purely for their sound, ignoring the meaning, to spell out Japanese words and grammar — the system now called 万葉仮名まんようがな (man'yougana). It worked, but it was heavy: a full character for every single syllable.

Both kana are what happened next, and they were made in two different ways. Hiragana came from writing those sound-characters in cursive until only a flowing outline was left. 安あん (an) melted into あ (a), and 加か (ka) into か (ka).

Katakana came from clipping one piece out of a character: the left side of 伊い (i) became イ (i), and the top of 多た (ta) became タ (ta). It started as shorthand jotted beside Chinese texts to help read them aloud in Japanese.

The family tree is that simple. Kanji came from China, and both kana were made out of kanji: one by melting, one by cutting.

Not really. ローマ字ローマじ (roomaji) is Japanese written in the Latin alphabet. It exists mostly for people who cannot read the other three: station signs for visitors, passports, and company names abroad.

Japanese adults do not read their own language in it.

It does live inside the keyboard, though. Japanese is typed by spelling the sounds in Latin letters and letting the software convert them. Type n-i-h-o-n and you get にほん (nihon), then convert once more to 日本にほん (nihon).

Romaji is a way in, not the writing itself.

For a learner it has one honest use — saying a word out loud before you can read kana — and one real cost. Every sentence you read in romaji is a rep your kana didn't get. The hiragana field note makes the case for dropping it in week one.

Hiragana first, and nothing else competes. It spells the grammar, so it appears in nearly every sentence you will read. Forty-six characters with almost no spelling traps is a realistic week of work.

The hiragana guide lays that week out day by day, with the look-alike shapes that cause most early confusion.

Katakana second. It is the same 46 sounds you already learned in different shapes, so you are learning forms rather than a new system.

It tends to feel harder than hiragana for one plain reason. It appears a few words per page instead of a few per sentence, so you get much less practice by accident.

The katakana guide covers the shapes people confuse and how English words get respelled.

Kanji third, and then for a long time. Japan's official everyday set is exactly 2,136 characters, which is the adult reading standard rather than an entry requirement.

The first hundred are the ones on price tags, station boards, and menus. The kanji guide explains why the order you meet them in matters more than the count.

One reassurance while you work through that order: you are never locked out of a sentence. A word whose kanji you have not met can be written in kana, sounded out, and looked up. That is the same fallback Japanese readers use, and it has a name (see the last question below).

A word that carries meaning takes kanji, and a word that shapes the sentence takes hiragana.

Katakana marks a word from outside, a sound effect, or a word set off for emphasis.

Each change of script marks a word boundary, which is why Japanese needs no spaces.

Three that Japanese people read every day: hiragana, katakana, and kanji. Romaji, the Latin alphabet, is sometimes counted as a fourth.

It serves people who cannot read the other three rather than being a script the language runs on. One ordinary sentence normally contains all three of the first three.

They're the same size and the same sounds, so neither is harder by design. Most learners find katakana slower to stick, because it appears far less often in ordinary text — fewer chance encounters, less free review. The fix is to read katakana on purpose: menus, drink shelves and product names are full of it.

For anything written for adults, yes. Kanji carry the content words, and a text without them gives you no word boundaries.

You do not need them to start, though. With kana alone you can read a children's book, sound out any word you meet, and type on a Japanese keyboard in your first week.

Japan's official everyday set is exactly 2,136 characters. Newspapers and novels stay inside that set almost entirely, with a tail of name and specialist characters outside it.

You can read a great deal with far fewer. The first few hundred cover what repeats everywhere: numbers, days, directions, places, and the most common verbs.

振り仮名ふりがな (furigana) is small hiragana printed above a kanji — or beside it, in vertical text — to give its reading. You'll see it in books for children, in material for learners, and over rare characters and unusual name readings in ordinary publishing. It's how a Japanese reader pronounces a character they've never met before.

Our kana pages teach the two syllabaries the way this article describes them: one sound system, two sets of shapes. Every character comes with stroke-order animation, a handwriting-style font, audio, and quizzes.

Hiragana and katakana sit next to each other, so the pairing stays visible while you learn. The kanji sets continue where the kana end, starting with the characters you will meet on a menu this week.

Japanese still has no alphabet. It has a sentence you can take apart, which turns out to be the more useful thing.