The Complete Katakana Guide
Katakana is the other kana script — sharper, more angular, and specialised. If hiragana is the everyday handwriting of Japanese, katakana is the neon sign. It's what you'll see on menus, in brand names, on imported products, in comic-book sound effects, and in a huge percentage of modern tech and scientific vocabulary. Learners sometimes treat it as optional because textbooks introduce it after hiragana, but if you can read hiragana and skip katakana you'll be locked out of reading roughly half of a typical Japanese street sign.
This guide covers everything: the script's origin, every base character, the voicing marks, the contracted sounds, the long-vowel dash that only appears in katakana, and the loanword rules that explain why "computer" ends up as コンピューター. By the end you'll know how to read a Japanese coffee menu without touching a dictionary.
What katakana is for
Modern Japanese uses four writing systems side by side: kanji (borrowed Chinese characters), hiragana (the flowing native syllabary), katakana (the angular native syllabary), and occasional rōmaji (the Latin alphabet). Each has a job. Katakana's jobs are:
- Foreign loanwords. Any word imported from another language after about 1600 is written in katakana. コーヒー (kōhī, coffee), テレビ (terebi, television), パン (pan, bread — from the Portuguese pão).
- Foreign names. Non-Japanese personal names and country names: アメリカ (Amerika), マリア (Maria), ブラジル (Burajiru, Brazil).
- Onomatopoeia. Comic-book sound effects and descriptive mimetic words: ドキドキ (doki-doki, heart pounding), ワンワン (wan-wan, a dog barking), キラキラ (kira-kira, sparkling).
- Scientific and technical terms. Animal and plant species names in scientific contexts, chemicals, and a lot of medical vocabulary: イヌ (inu, "dog" written in katakana for a biology textbook), アルミニウム (aluminium).
- Emphasis. Sometimes used the way English uses italics or bold — to draw attention to a word.
Origin: the monks' shorthand
Like hiragana, katakana descends from kanji — but the path is different. While hiragana came from cursive simplifications, katakana came from fragments of kanji, taken as abbreviations. Buddhist monks in the 9th and 10th centuries annotated Chinese religious texts with small marks to indicate Japanese pronunciation, and rather than write a full kanji in a cramped margin they'd take a single component. Over time those components crystallised into the standard script.
You can still see the kanji origins if you look carefully. ア (a) comes from the left half of 阿; イ (i) from the left radical of 伊; カ (ka) from the top half of 加; サ (sa) from the top of 散. Because each katakana is a piece of a kanji rather than a cursive rendering, the script feels harder and more geometric.
The 46 base characters
Katakana mirrors hiragana one-to-one: same rows, same syllables, same order. If you already know hiragana, you already know the sound each katakana makes — you just need to learn the new shape.
- ア イ ウ エ オ — a, i, u, e, o
- カ キ ク ケ コ — ka, ki, ku, ke, ko
- サ シ ス セ ソ — sa, shi, su, se, so
- タ チ ツ テ ト — ta, chi, tsu, te, to
- ナ ニ ヌ ネ ノ — na, ni, nu, ne, no
- ハ ヒ フ ヘ ホ — ha, hi, fu, he, ho
- マ ミ ム メ モ — ma, mi, mu, me, mo
- ヤ ユ ヨ — ya, yu, yo
- ラ リ ル レ ロ — ra, ri, ru, re, ro
- ワ ヲ ン — wa, wo, n
Dakuten and handakuten work identically to hiragana: カ → ガ (ka → ga), サ → ザ (sa → za), ハ → バ → パ (ha → ba → pa). Yōon combos use a small ャ ュ ョ: キャ (kya), シュ (shu), チョ (cho).
The long-vowel dash ー
Here's the one rule that genuinely differs between the two scripts. Katakana indicates long vowels with a horizontal dash: ー. It doesn't matter which vowel; the dash just extends whatever came before it.
- コーヒー (kōhī, coffee) — both o and i extended by the dash.
- ケーキ (kēki, cake).
- スーパー (sūpā, supermarket).
- ニュース (nyūsu, news).
The dash rotates to vertical (a stroke like ｜) in vertically-typeset Japanese, which you'll see on book spines and some newspaper columns. Don't be thrown by it — same character, same function.
The extra sounds katakana needs
Loanwords brought sounds that native Japanese doesn't have: v, ti, di, she, je, fa, fi, fe, fo. To spell these, katakana uses small-vowel combos that hiragana doesn't need.
- ヴ (vu) — used for V sounds: ヴァイオリン (vaiorin, violin). Many writers use バ instead.
- ティ (ti) — パーティー (pātī, party).
- ディ (di) — ディズニー (Dizunī, Disney).
- シェ (she) — シェフ (shefu, chef).
- ジェ (je) — ジェット (jetto, jet).
- ファ フィ フェ フォ (fa, fi, fe, fo) — ファックス (fakkusu, fax), フォーク (fōku, fork).
- ウィ ウェ ウォ (wi, we, wo) — ウィーン (Wīn, Vienna).
You don't need to memorise these up front — they pop up naturally as you read loanwords. Just know that katakana is a more flexible system than hiragana for exactly this reason.
Loanword rules you can crack
Japanese syllables are built from consonant + vowel. When a foreign word arrives with consonant clusters or final consonants, Japanese inserts vowels to break them up. The default vowel is u, except after t and d (where it's o). So "strike" becomes ストライク (sutoraiku): s-u-t-o-r-aiku. Once you see the pattern you can read loanwords you've never seen before just by sounding them out.
- Final L or R is usually dropped or becomes ル: hotel → ホテル (hoteru).
- Final consonants get a vowel added: bus → バス (basu), test → テスト (tesuto).
- Clusters are broken: Christmas → クリスマス (kurisumasu), stress → ストレス (sutoresu).
- V sounds often become B: video → ビデオ (bideo).
- Th sounds become S or Z: thank you → サンキュー (sankyū).
- R and L both become the Japanese R: rice and lice both become ライス (raisu). Context sorts it out.
The lookalikes that trip everyone up
Katakana's biggest challenge is visual similarity. A few pairs are almost identical and cause massive confusion in the first week:
- シ (shi) vs ツ (tsu) — both have three strokes arranged similarly. The key: シ has strokes entering from the left flowing to the bottom-right; ツ has strokes entering from the top flowing to the bottom-left. Stroke direction is the giveaway, not position.
- ン (n) vs ソ (so) — the same stroke-direction trick: ン's dot is near the bottom and the curve travels upward; ソ's dot is near the top and curves downward.
- コ (ko) vs ユ (yu) — ko is two horizontal strokes opening right; yu adds a vertical that crosses the top.
- ク (ku) vs ワ (wa) vs フ (fu) — all similar hook shapes. Ku has a clear angle change; wa is rounder with a longer curve; fu is the flattest, almost a single gentle hook.
- ル (ru) vs レ (re) — ru has a small hook on the right; re is a plain L.
Full side-by-side breakdowns live in our confused-pairs guide. The confused-pair drill specifically targets these using spaced repetition on the pairs you get wrong.
A katakana reading sample
Try this menu excerpt: コーヒー、カフェラテ、ホットチョコレート、オレンジジュース、パンケーキ、サンドイッチ、サラダ. Did you get kōhī (coffee), kafe rate (café latte), hotto chokorēto (hot chocolate), orenji jūsu (orange juice), pankēki (pancakes), sandoitchi (sandwich), sarada (salad)? If any of those took you longer than three seconds, that's the area to focus your next drill session.
How to practise
The practical loop is identical to hiragana but the content is different: more loanword examples, more long-vowel dashes, more extended-sound combos. Our recommendation:
- Finish hiragana first. Don't mix them — your brain will blur the visually similar shapes across scripts.
- Use the guided path set to katakana; it introduces one row at a time with shared-sound hints from the hiragana you've already learned.
- Focus early practice on the confused-pair drill — シ/ツ and ン/ソ are the confusions that derail beginners.
- Read loanword-heavy material for real-world application: a café menu, a tech product page, or a pop music track list.