I've been casually studying Itak on and off in between my usual schoolwork and obvious Japanese studies. However, I've been more and more bothered by a particular issue as I've gone on: how does one go about romanizing Itak from standard katakana?

I'm aware that this site boasts a small guide, but that doesn't really work too well when the dictionary you're using doesn't follow those rules. I'm not sure if a standard guide even exists, but less how to go about transliterating this language to English characters. I've seen odd ones where people will write a name like "cup=kamuy", when I'll see others write as "Chupkamui". There are probably a ton of variations like with Japanese as well, but I lack the understanding of which ones are standard, which are more widely understood, and which are better to use overall.

Any and all advice one has on this will be taken into account and greatly appreciated. If further clarification is needed, feel free to ask. Thank you for reading this.

  • omniglot.com/writing/ainu.htm <---That's a better guide that's I've found, but it's still rather inconsistent in places.
    – Pleiades
    Commented Oct 25, 2016 at 20:11
  • 1
    A corpus of Ainu Forklore run by the National Institute for Japanese Language and Linguistics describes its rule for Latin transcriptions. I am not sure it's the standard or not, but might be a valuable reference.
    – Yosh
    Commented Oct 27, 2016 at 4:01
  • This might be one of the better ones I've either found myself or been given over the last few days. Definitely a valuable resource.
    – Pleiades
    Commented Oct 27, 2016 at 4:11
  • 1
    You might also find Batchelor's work useful. It's a bit outdated, and the orthography has shifted a bit (Batchelor doesn't use the small kana that are currently standard for Ainu), but it's still a decent resource. Commented Feb 2, 2017 at 20:59

2 Answers 2


I'm not sure that there is a good standard, but it may or may not matter, given the relatively tiny audience there is for any such transliterated material. Anna Bugaeva is one person who's written a good deal of scientific papers in English about Aynu, and you might use her transliteration as a starting point.

Personally, I just transcribe it however seems best to me, which generally involves making it look as much like a language people actually use as I can - I throw out the equals sign entirely (since anywhere else that's just a technical notation for clitic attachment), and I tend to use ch and y for others' c and i. I'd love to see a kanji/hiragana orthography for Aynu, since the katakana one looks so much more like a scientific transcription than a language someone actually speaks, but that's not an easy thing to do.

In short, figure out how it actually sounds and romanise it however you like, but feel free to use the existing scientific literature as a guide if you want.


The de facto standard method for transcribing Ainu (both in Latin alphabet and in katakana) being used today is the one proposed in Akor Itak, a textbook published by the Hokkaido Utari Association (now Hokkaido Ainu Association) in 1994. It is often referred to as “(nearly) phonemic transcription” ((簡易)音素表記 in Japanese). What makes it different from some of the older romanization methods (e.g. the one devised by Batchelor in his Ainu-English-Japanese Dictionary) is the usage of ⟨c⟩, ⟨s⟩, ⟨y⟩, ⟨w⟩, where others would use ⟨ch⟩, ⟨sh⟩, ⟨i⟩, ⟨u⟩ respectively (the last two rules only apply in syllable-final position, after a vowel). For example, the words transcribed as “ikkeu” and “kamui-chep” by Batchelor are written as “ikkew” and “kamuycep” in modern dictionaries. Apart from that, the equal sign “=” is used to indicate personal affixes (e.g. “ci=nukar”, rather than “chinukar”). For more examples, you may refer to the following resources (all of them apply the romanization standard described above):

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .