Katakanization is the transliteration of foreign words to katakana, i.e., their writing using the closest-sounding katakana characters. We katakanize foreign words when there are no original Japanese words for them, e.g., there is no Japanese word for ‘piano’, so we always katakanize it, and we will always find it as ピアノ (piano).

However, even when there is a Japanese word for the foreign word, we may still want to katakanize it, to preserve an original name, title, or quote. For example, the Japanese word for ‘bicycle’ is ‘jitensha’, so normally we would write it in kanji (自転車) or in hiragana (じてんしゃ); in some cases, say, a movie title with the word ‘bicycle’ in it will be translated to Japanese; for example, the Italian movie ‘The Bicycle Thieves’ was translated as じてんしゃどろぼう (jitenshadorobau). However, in other cases the title is not translated but transliterated, i.e., katakanized; for example, the title of Queen’s single ‘Bicycle Race’ was katakanized as バイシクル・レース (baishikuru reesu). Thus, in general, katakanization can happen to any foreign-origin word, regardless of whether there is already a Japanese word for it or not.

Katakanization influences

The Japanese culture excels at borrowing, adapting and improving over many things other cultures have to offer. It might be the tea and ramen from China, or the piano and violin from the western musical tradition, or the baseball and burger joints from the United States. The Japanese written language also has borrowed heavily from other cultures, specially from China, from which it borrowed the kanjis multiple times. See the interesting Japanese: The borrower language post, in Tofugu, for more details on the external influences of Japanese.

Japanese words borrowed from a language other than Chinese are called ‘gairai-go‘ (外来語), or ‘imported language’. In a few cases, we may use kanjis to write gairai-go, e.g.,

Japanese / romaji
合羽 / kappa
天ぷら / tempura
英吉利 / igirisu

from / origin
capa de lluvia / Spanish
têmpora / Portuguese
inglez / Portuguese


And there are a few words that are half-gairaigo, so we write the Japanese part in kanji or hiragana, and the foreign part in katakana, e.g.,

half-gairaigo / romaji
ビー玉 / bii-dama
歯ブラシ / ha-burashi
消しゴム / keshi-gomu
シャボン玉 / shabon-dama

from / origin
vi(drio)-ball / Port-Japan
tooth-brush / Japan-Eng
vanish-gum / Japan-Dutch
jabon-ball / Spanish-Japan

glass-ball, marble
soap bubble

However, in general, we write gairai-go words in katakana. Usually, words of Chinese origin are not considered gaira-go, and we write them with kanjis instead of katakana. However, there are exceptions, like ‘ramen’, of Chinese origin, that we commonly find katakanized as ラーメン.

The ichiraku ramen shop from the Naruto series

Some examples of gairai-go are:


part-time work


Still, in terms of impact, after Chinese, the language that has influenced Japanese the most is English, which happened to be a lingua franca in the past century of globalization. Baseball, of American origin, is widely popular in Japan, to the point that it has its own Japanese word, with its own kanji version: ‘ya-kyuu’, or 野球 (field-ball). Likewise, fast food (ファストフード, ‘fasutofuudo’) and convenience stores (コンビニ, ‘combini’) are ubiquitous, blue-jeans (ブルージーンズ, ‘buruujiinsu’) and t-shirts (Tシャツ, ‘tiishatsu’) are as common as in Texas, and the language is peppered with anglicisms, popping anywhere from anime songs to everyday expressions like バイバイ(‘baibai’, bye-bye), and ドンマイ (‘donmai’, from “don’t mind” meaning “don’t worry about it”).

‘donmai’ (ドンマイ, ‘don’t mind’) and ‘imeeji’ (イメージ, ‘image’), from ‘Horimiya’

We can trace around 10% of all Japanese words to a closely-sounding English word. Thus, if we don’t know how to say an English word in Japanese, like ‘milkshake’ or ‘handbell’, we can just katakanize it and we’d have a 10% chance of getting it right! Indeed, ‘milkshake‘ is ミルクセーキ and ‘handbell‘ is ハンドベル. This works with a lot of words, even everyday ones like ‘ball’ (ボール) or, as shown above, ‘image’ (イメージ), meaning ‘mental image’ or ‘impression’, and it works particularly well with places, food, clothes, and technological and scientific words of western origin.

What word to katakanize?

Let’s put ourselves in the role of a Japanese. How would we katakanaize a word? It seems like a straight forward process. There are two issues, though. The first one is deciding what word we should katakanize. Let’s consider the case of katakanizing the word ‘ladies’. Since Japanese doesn’t have plurals, we have to katakanize the singular ‘lady’ instead, as レディー (redii), or レディ (redi). However, we need to keep the plural ‘ladies’ in some contexts, e.g., “ladies’ fashion” is レディースファッション (rediisu fasshen). Also, are we katakanizing a word from its original language, like ‘Deutschland’, or its English version, like ‘Germany’? Although with different popularity, both ‘Deutschland’ as ドイチュラント (doichuranto), and ‘Germany’ as ジャーマニー (jaamanii) are used.

The second issue of katakanizing a word is considering how it is pronounced. Are we katakanizing the British pronunciation, traditionally used, or the American one, more influential in the past few decades? Take, for example, the word ‘sugar’; it’s katakana, based on the British pronunciation, would be シュガー (shugaa), but based on the American pronounciation, it would be ‘shugaru’. It turns out that ‘sugar’ is normally katakanized using the British version.

If we were katakanizing words from a regular language, like Spanish or German, then it would be possible to automate this step because the vowels and consonants would tend to have a unique sound. For example, we katakanize ‘Buenos Aires’ as ブエノスアイレス (buenosu airesu), ‘enchilada’ as エンチラーダ (enchiraada), and ‘Deutschland’ as ドイチュランド (doichurando); each Spanish or German letter and letter combination maps to a predictable sound or sounds, so mapping them to Japanese is easy. Mapping English to Japanese is not this straightforward because English is very irregular; there is no rule to tell us that the ‘s’ in ‘sugar’ sounds ‘sh’ to get ‘shugaa’, or that the ‘s’ in ‘island’ is silent, to get アイランド (‘airando’).

There is a similarity bewteen English and Japanese about this uncertainty of how a character is pronounced. In Japanese, a single kanji can have many sounds, so without context we cannot pronounce it; hence, a combination of kanjis is just a reminder of how the word sounds. Likewise, in English, a single letter can have multiple sounds, so without context we cannot pronounce it; hence, a combination of English characters is just a reminder of how the word sounds. In the same way that there is no rule to predict what a single out-of-context kanji in Japanese sounds, there is no rule to predict how a single out-of-context character in English sounds.

Luckily for us, there is a large flexibility on how English words are actually perceived. For example, does ‘diode’ sound ダイオード (daiood) or ダイオド (daiod)? hum… hard to tell. But… why fight? Japanese accepts both as possible ways to write ‘diode’. The same happens with a large number of katakana words. Another example is the katakanization of the ‘v’ using either a ‘v’ or a ‘b’, e.g., we can write ‘violin’ as バイオリン (baiorin), ヴァイオリン (vaiorin), ヴィオロン (vioron), or ビオロン (bioron). As long as the context makes clear what we are talking about, there is room to alter the katakana and still get it right.

In summary, there are so many cases and exceptions of how to pronounce an arbitrary English word that coming up with a good automatic katakanizer looks like a daunting proposition; still, we can find an attempt here.

from katakana back to English

For the reverse process of figuring the foreign word that originated a katakana, we will assume that the katakana was transliterated from English; if it wasn’t, the process might fail. For example, katakanizing the English word ‘centaur’ would give us センタアー (sentaaa), but this is incorrect; instead, the Japanese word for ‘centaur’ is ケンタウロス (kentaurosu) because they katakanized the original Greek word ‘kentauros’. Still, as English speakers we are in luck because English is, by far, the main source of katakanized words, so assuming that a katakana was transliterated from an English word will lead us to the correct answer most of the time.

Suppose we ran across the katakana word セラー, i.e., ‘seraa’. If ‘seraa’ was derived from English, then ‘seraa’ is the sound of the word as heard by the Japanese person. If we understand how a katakana was made, we should be able to tell that the likely English words that this person could have katakanized as ‘seraa’ are [sc/c/s]e[r/rr/l/ll][ar/er/or]; all these posible combinations are

scerar scerrar scelar scellar cerar cerrar celar cellar serar serrar selar sellar
scerer scerrer sceler sceller cerer cerrer celer celler serer serrer seler seller
sceror scerror scelor scellor ceror cerror celor cellor seror serror selor sellor

Of these words, only ‘cellar’ and ‘seller’ are actual English words and those two words are, indeed, the only two possible meanings of the katakana セラー.

how to katakanize an English word

How to katakanize anything in one sentence:

Write in katakana the sounds of the words as a Japanese person would hear them.

That’s all. In a few cases, we need to take into account the writing, e.g., “media” is メディア, as it is written instead of as it sounds. However, by and large, katakanization depends on the sound of the word.

We are more likely to succeed if we katakanize the British pronunciation; on this token, here is some basic katakanization of the British ‘r’:

word-end ‘vowel + r’ → aa

[a/u]r → aa

or → oo

car (kaa), power (pawaa), sir (saa), color (karaa), fur (faa)

army (aamii), artist (aatisuto),
urchin (aachin), urgent (aashento)

order (oodaa), force (foosu),
formal (foomaru), mortal (mootaru)

If the katakanization fails and the word is recent, we should try the American pronunciation.