How to learn Japanese by reading

Japanese has a reputation as one of the hardest languages in the world to read. Three scripts share every page, no spaces separate the words, and most kanji have more than one reading. All of that is real. But Japanese has things going for it too. The grammar is remarkably regular, and Japanese teachers even have a name for learning by reading a lot of easy text: tadoku (多読), literally “much reading.”

Once the scripts are familiar, reading is how the rest falls into place: the kanji you meet in words, the grammar you meet in sentences, and the rhythm of a language in which the verb always waits until the end.

Three scripts, no spaces

A single sentence mixes all three: 白いウサギが時計を見た。“The white rabbit looked at its watch.”

• Hiragana writes grammar, such as particles and verb endings, and many native words: が, を, た.
• Katakana writes loanwords and foreign names, and often animals and sound effects: ウサギ (usagi, rabbit), コーヒー (kōhī, coffee).
• Kanji write most content words: 白 (white), 時計 (watch, clock), 見 (see).

Because there are no spaces, the change from kanji to kana is your main clue to where words begin and end. That’s why text written mostly in kana, though easier to pronounce, can be harder to read: a long run of hiragana has no landmarks. Books for young children often add spaces between phrases; adult text doesn’t.

Then come the kanji themselves. The official list of characters for general use, the jōyō kanji, has 2,136 entries, and most have at least two kinds of reading: a Sino-Japanese on reading and a native kun reading. 生 is sei in 学生 (gakusei, student), i in 生きる (ikiru, to live) and u in 生まれる (umareru, to be born), among others. You don’t learn readings in isolation. You learn words, and the words teach you the readings. Our guide to learning a new script has more on getting started with kana.

The grammar you meet in stories

• Particles mark roles. は (wa) marks the topic, が (ga) the subject and を (o) the object; に marks time, destination and where something is, and で where an action happens and the means. The difference between は and が is subtle, and it’s learned best from many examples.
• The verb comes last, and everything that describes a noun comes before it: 昨日駅で会った人 is “the person (I) met at the station yesterday,” literally “yesterday station-at met person.”
• Subjects disappear. Japanese drops “I,” “you,” “he” and “she” whenever context makes them clear, which in a novel is most of the time. You follow who’s doing what through verb forms, politeness and the flow of the scene.
• Two registers. Polite -masu forms (行きます) and plain forms (行く) mean the same thing but carry different social weight. Most novels narrate in the plain form, and the way characters talk tells you about them: an old professor, a schoolgirl and a gangster each sound different on the page.
• Sound words. Japanese has a rich stock of onomatopoeia and mimetic words, such as ドキドキ for a pounding heart and ニヤリ for a sly grin. Fiction is full of them.

Japanese reading at each level

What the CEFR levels look like in Japanese text (see CEFR levels explained for the scale itself):

• A2, elementary. Short sentences, mostly in kana with basic kanji, often in the polite -masu form in material for learners. At A2 and B1, Lexo prefers kana for kanji above roughly the JLPT N3 level, so you aren’t stopped by characters you haven’t met yet.
• B1, intermediate. Plain-form narration, the te-form linking one action to the next, clauses that describe nouns, and kanji up to an intermediate level.
• B2, upper intermediate. Kanji as in ordinary books, passive and causative verbs (〜られる, 〜させる), several kinds of conditional (〜ば, 〜たら, 〜なら), and the written style that links clauses with a bare verb stem (取り出し、…) rather than the te-form.
• C1, advanced. Close to the original, lightly condensed, with literary vocabulary and the full range of politeness in dialogue. C2 is the complete text.

Alice at A2 and at B2

The opening of Alice’s Adventures in Wonderland, written for this page at two levels:

A2

アリスは白いウサギを見ました。ウサギはポケットから時計を出しました。それから、あなにとびこみました。アリスもとびこみました。

Alice saw a white rabbit. The rabbit took a watch out of its pocket. Then it jumped into a hole. Alice jumped in too.

B2

退屈していたアリスは、チョッキのポケットから懐中時計を取り出し、「たいへんだ、遅刻しそうだ」とつぶやきながら走っていく白ウサギを見かけて、思わずそのあとを追いかけた。

Alice, who had been bored, spotted a White Rabbit taking a pocket watch out of its waistcoat pocket and hurrying past, muttering, “Oh dear, I’m going to be late,” and without thinking she ran after it.

The A2 version is four short polite sentences. Everyday kanji like 白, 見 and 時計 stay as kanji, while あな (“hole”) and とびこみました (“jumped in”) are spelled in kana, as a young reader would see them. The third sentence also drops its subject: nothing says “the rabbit,” and context does the work. The B2 version is one plain-form sentence, and everything from チョッキ to 走っていく is a single long clause describing the rabbit. The verbs 見かけて and 追いかけた arrive only at the end. Reading Japanese at this level means holding that long description in mind until you reach the noun it belongs to.

Building a Japanese reading habit

1. Learn kana first, completely. Before anything else, learn hiragana and katakana until you can read them without thinking. Many learners need a few weeks.
2. Read first, then listen. Read a section on your own, then play it with the text highlighted as it’s read. Hearing each kanji compound pronounced in context fixes its reading better than any list.
3. Tap words, not characters. When a word stops you, tap it for its meaning in context and play it to hear how it’s read. Don’t stop for every kanji you couldn’t have written yourself.
4. Reread yesterday’s pages. Rereading in Japanese is fast and satisfying: the kanji you struggled with yesterday become familiar faces.

Choose the level carefully. According to Hu and Nation (2000, “Unknown vocabulary density and reading comprehension”), most learners need to know about 98 percent of the words in a story to understand it adequately without help. In Japanese, an unknown word written in kanji often can’t even be sounded out, so text that’s slightly too hard quickly becomes text you can’t read. If you’re tapping something in every line, drop a level. The 98% rule explains the research, and extensive reading and reading while listening cover the habits.

Classics to start with

• Alice’s Adventures in Wonderland has short chapters, simple settings and plenty of dialogue. A good first book at A2 or B1.
• Anne of Green Gables has long been a favorite with Japanese readers, and Anne’s torrent of talk is great practice at B1 and B2.
• The Adventures of Sherlock Holmes is a collection of short stories, each complete in a few sittings, at B1 and B2.
• Journey to the Center of the Earth has three main characters and one direction, down, which makes it easy to follow at A2.
• The Time Machine is short and thought-provoking, a good B2 read.

How Lexo helps you read Japanese

Lexo eases the kanji problem in two ways. At A2 and B1 it prefers kana for kanji above roughly the JLPT N3 level, so early reading isn’t a kanji hunt; from B2 up, kanji appear as they would in an ordinary book. And when a kanji compound stops you, the word card gives its reading in romaji, and the word itself can be played on its own, so you can see and hear how it’s read.

You can tap any word, even though nothing separates the words on the page. The card shows what the word means in that sentence, its base form and its part of speech (tap 見かけて and you get 見かける), explains any expression it belongs to, and translates the whole sentence into English. Keep the useful ones in your Words list.

The whole book is narrated, with different voices for different characters, and the text highlights as it’s read, tying each kanji to its sound. Sentence-by-sentence mode shows one sentence at a time with its audio and translation, and you can enlarge the type when the strokes of a kanji get hard to make out.

Pick the level per book, from A2 to C2. If no one has asked for a book in Japanese at your level yet, you’ll be the first: Lexo writes the edition, chapter 1 arrives in a few minutes, and the rest keeps coming while you read. First readers keep a dated “First edition” mark. There’s no account to create, and books stay on your iPhone or iPad for offline reading.

Kanji and Chinese characters share their roots; see Chinese, and for a grammar that works much like Japanese, Korean.

Lexo is coming soon to the App Store for iPhone and iPad. Get notified at launch.