Katakana Table
| Char | Rōmaji | From (kanji fragment) | Note |
|---|---|---|---|
| ア | a | 阿, left side | |
| イ | i | 伊, left side | |
| ウ | u | 宇, top | |
| エ | e | 江, right side | |
| オ | o | 於, left side | |
| カ | ka | 加, left side | |
| キ | ki | 幾, left side | |
| ク | ku | 久, left side | |
| ケ | ke | 介 | |
| コ | ko | 己, top | |
| サ | sa | 散, top | |
| シ | shi | 之 | strokes run horizontal; ツ runs vertical |
| ス | su | 須, top | |
| セ | se | 世 | |
| ソ | so | 曽, top | tick sits upper-left; ン lower-right |
| タ | ta | 多, bottom | |
| チ | chi | 千 | chi, not ti |
| ツ | tsu | 川 | strokes run vertical; シ runs horizontal |
| テ | te | 天, top | |
| ト | to | 止, top | |
| ナ | na | 奈, left side | |
| ニ | ni | 仁, left side | |
| ヌ | nu | 奴, right side | |
| ネ | ne | 祢, left side | |
| ノ | no | 乃 | |
| ハ | ha | 八 | |
| ヒ | hi | 比 | |
| フ | fu | 不, top | a single soft stroke |
| ヘ | he | 部, right side | same shape as hiragana へ |
| ホ | ho | 保, bottom | |
| マ | ma | 末, bottom | |
| ミ | mi | 三 | |
| ム | mu | 牟, top | |
| メ | me | 女 | |
| モ | mo | 毛 | |
| ヤ | ya | 也 | |
| ユ | yu | 由 | |
| ヨ | yo | 与 | |
| ラ | ra | 良, right side | |
| リ | ri | 利, right side | |
| ル | ru | 流, bottom | |
| レ | re | 礼, right side | |
| ロ | ro | 呂 | |
| ワ | wa | 和 | |
| ヲ | wo | 乎 | archaic; the object particle relic |
| ン | n | late invention | tick sits lower-right; ソ upper-left |
| Gadget | What it does | Example |
|---|---|---|
| ー | lengthens the vowel one beat | コーヒー = koohii, coffee |
| small ッ | doubles the consonant that follows | サッカー = sakkaa |
| small ゃゅょ | glue a y-sound onto i-column kana | シャ = sha |
| ヴ | the rare voiced u for foreign v | ヴィーナス = Venus |
| ファ ティ | loanword combos the grid never had | ファミリー, パーティー |
Katakana is the same 46 sounds as hiragana in work clothes: angular shapes chopped from fragments of Chinese characters - a corner of 安 became ア, a sliver of 加 became カ - by monks who needed a shorthand for reading Buddhist texts aloud. Today the script owns the foreign: loanwords, brand names, country names, sci-fi robots and menu items all default to katakana, which is why every product label in Japan reads half-English through this chart.
Two mechanical add-ons do most of the heavy lifting. The long-vowel bar ー (chōonpu) stretches the previous vowel one beat - コーヒー is koohii, four beats, not kohii - and the small ッ doubles the consonant that follows, the same beat-doubling hiragana writes with small っ. Beyond the 46, loanword pressure has forced combos the grid never had: ファ (fa), ヴィ (vi), ティ (ti) - small vowels glued to extras, which is why katakana keeps growing while hiragana never does.
How to use
- Map each katakana onto its hiragana twin: ア イ ウ エ オ pair with あ いうえお one-to-one, so learn this chart as the second costume of sounds you already know, not a new alphabet.
- Sort the lookalikes by stroke logic: シ and ツ differ in stroke direction (horizontal vs vertical), ソ and ン in which corner the tick sits, and ナ meets メ only at a glance - drill the pairs, not the singles.
- Read loanwords by beats: count one beat per kana plus one per ー bar and one per small ッ, because Japanese rhythm is syllable-timed - ビール (biiru) is three beats, and miscounting is the difference between beer and building (ビル).
Frequently asked questions
How do I tell シ and ツ apart?
Follow the strokes, not the silhouette: シ (shi) is built with two small strokes on the left and a long sweep underneath - they write horizontally, left to right; ツ (tsu) stacks the same three strokes vertically, top to bottom. The near-foolproof test: find the two small strokes and ask which way the line of them runs - horizontal means shi, vertical means tsu. Font styles blur this (some typefaces tilt everything), which is exactly why the loanword script gets a reputation for unreadability that hiragana escapes.
Why does Japanese use katakana for foreign words?
Role assignment, not difficulty: after the two kana scripts settled, katakana became the margin script - used for annotations, official readings and anything marked as not-native. Foreign words fell into that slot naturally, and modern usage keeps the logic: loanwords (テーブル, teeburu), foreign names (ロンドン, Rondon), countries, species names, onomatopoeia and robotic shouting all wear it. The practical payoff for learners is searching - if a word is English in origin, typing it in katakana into a map or menu finder usually works, while hiragana would look like a grammar word.
What is the ー bar and where did it come from?
The chōonpu (長音符), a lengthening bar that adds one vowel beat to whatever kana precedes it: ラーメン is raa-men (four beats), ケーキ kee-ki. It is essentially katakana-only - hiragana writes long vowels the honest way, by spelling them out with a second kana (らあめん), and kanji ignores the concept. The bar exists because loanwords carried vowel lengths Japanese script had no letters for; a line was the cheapest invention that fit the angular grid. Misreading it compresses words - ビール loses its second beat and drifts toward ビル, building.
What are combos like ファ and ヴィ?
Loanword pressure patches. The Japanese grid has no fa, fi, ti, di, vu or ye - so modern spelling glues a small vowel to an existing kana: ファ (fa) is fu plus small a, ティ (ti) is te plus small i, ヴィ (vi) uses the rare voiced-u ヴ. Purists resist (official style often rewrites them as フua or fu for foreign bodies), but menus, brands and names use them daily. The list keeps growing - フェ fye, ウィ wi - which is why katakana charts in textbooks lag behind real signage.
What does the small ッ do - and how is it different from a full-size つ?
The small ッ doubles the consonant that follows it, adding one silent beat before it arrives: サッカー is sak-kaa (four beats), きって is kit-te. The full-size つ is an ordinary letter with its own tsu sound, so size is meaning - the same trick hiragana writes with small っ (きって again). In loanwords the ッ usually marks where the foreign word had a double consonant (ベッド, beddo, from bed), and getting it wrong changes words: おばさん (obasan, aunt) versus おばあさん (obaasan, grandmother) differ only in beats.