I had the impression that linguists these days generally consider there to be a Japonic language family consisting of Standard Japanese, the Ryukyuan languages including Okinawan, and Hachijo, spoken on another group of islands south of Tokyo.

Many or all of these languages in turn have their own dialects. The distinction between dialect and language being the usual one in linguistics based on whether any two varieties are mutually intelligible.

But then it was pointed out to me that the prominent Japanese linguist Masayoshi Shibatani, author of The Languages of Japan regards these all to be dialects of Japanese. The book was published fairly recently, 1990.

Some years ago ISO added language codes for a number of these. UNESCO also recognized them as endangered languages.

But do we know if there is a consensus among linguists on either side?

  • 3
    "The distinction between dialect and language being the usual one in linguistics based on whether any two varieties are mutually intelligible". This definition has inherent problems. Mutual intelligibility is not a transitive relation, but "being the same language" is.
    – dainichi
    Commented Mar 13, 2014 at 1:35
  • Indeed. I didn't want to try to cram in everything about dialects into my question, there is not really a "natural" distinction anyway, just ways of looking at things and analysing them in more or less useful ways. Commented Mar 13, 2014 at 19:41

3 Answers 3


Having talked to Shibatani after the book was published, I can pretty safely say that he has since changed his mind.

I think the basic consensus among linguists who have some specialization in Ryukyuan is that there are at least seven Japonic languages in two subgroups. The first subgroup is the Japanese subgroup, which includes Japanese proper, as well as Hachijo. The second subgroup, Ryukyuan, includes at least Amami, Okinawan, Miyako, Yaeyama, and Yonaguni. Amami and Okinawan may be a dialect chain (the division between the two is not well studied), and it's almost certain that there are more than just three languages (Miyako, Yaeyama, and Yonaguni) in the southern Ryukyus. The proposal for the forthcoming Handbook of the Ryukyuan Languages reflects the consensus view I outlined above (minus Hachijo, which is not a part of the Ryukyuan subgroup).

The ISO codes make absolutely no sense. Nobody that I know of, and no source that I've ever seen groups the languages like they do. Unfortunately, this has trickled down into Ethnologue, which has its own problems in dealing with the Japonic languages. For instance, they claim that there are 984,000 speakers of Okinawan. While the Japanese government does not track language use data on their census, this number is far too high--first and foremost, it includes more people than actually live in the area where their definition of Okinawan is spoken. The actual number of speakers is likely less than 150,000 speakers.

  • 1
    A great answer. Interesting that it is ISO influencing Ethnologue. Until now I had always had the impression that it was Ethnologue influencing ISO. But I don't know very much about how either work other than Ethnologue being related to missionary work. Commented Mar 13, 2014 at 10:40
  • It might have been the other way around. I haven't dug into the history of the proposal. But like I said, the ISO codes need to be fixed at some point.
    – limetom
    Commented Mar 14, 2014 at 2:12
  • They do revise the ISO language codes every few years, I'm not sure whether it's in step with each new edition of Ethnologue. There are ISO language code mailing lists which are at least readable by the public. I've stumbled over them while Googling several times and there's lots of interesting discussion, disagreement, etc in them. Commented Mar 14, 2014 at 13:10

The problem with the language/dialect distinction is that it always, always, always involves politics. Even when trying to do things on a purely linguistic basis, there are far too many borderline cases and questionable situations (e.g. dialect continua, one-sided intelligibility, etc) - you will never end up with a wholly unbiased classification.

That said, most (maybe even all) non-Japanese linguists recognise Ryuukyuuan as a group of languages separate from Japanese, regardless of how many Ryuukyuuan languages they count. I would guess that the 'canonical' Japanese view, that Ryuukyuuan is just a set of Japanese dialects, comes from nationalism and the very inaccurate and outdated view of Japan as an almost totally homogenous country - even Ainu was called a 'dialect', and it's totally unrelated to Japanese! I don't know how many Japanese linguists still regard them as dialects instead of languages, though.

The linguistic criteria fairly unambiguously support the separate-language analysis - there is a clear gap in intelligibility between the Kyuushuu mainland and the Amami archipelago. It actually turns out that Ryuukyuuan as a whole is genetically closer to Northeast Kyuushuu than anywhere in the south, but it took until 2003 (in a paper by Leon Serafim) for anyone to notice. If you show Japanese speakers a video of someone speaking Ryuukyuuan (without explaining it) and ask them if it's Japanese, they will definitely say it's not - I actually came across a video on YouTube recently that was a quiz asking the viewer to distinguish audio clips of Japanese dialects and Ryuukyuuan languages from clips other East Asian languages :P

Now, it may turn out that Ryuukyuuan is genetically a sub-branch of a Japanese-internal dialect group, but it has been innovative enough since it split off that it still counts as a group of separate languages.

I don't know about Hachijou - I know it's very very different from the rest of Japanese, but I don't know if there's a consensus as to a separate status. If I were to venture a guess, I would bet that not only Hachijou but also a number of other dialects from Japan's main islands ought to be considered separate languages rather than dialects - though these might be a bit more muddled by the dialect continuum problem than Ryuukyuuan or Hachijou.

  • Not just Ainu, but Palauan was at one point called a dialect of Japanese during the highest extent of Japanese colonialism.
    – limetom
    Commented Mar 13, 2014 at 8:39

I think you are just using old stereotype of Japanese as nationalist/racist which in itself is a prejudice.

For one, if unintelligibility is the drawing line, then there are about 20 languages in japan, one of which is going to be ryukyuan.

When you said "most (maybe even all) non-Japanese linguists" you probably just referring to those who are familiar with Western convention. If one take Japanese or Chinese convention, Dutch/German or Spanish/Italian are just dialect of Germanic language and Spanish and Italian are dialect of Roman language. This definition make far more sense, if these lanauges are written in characters instead of alphabet. Also, the equivalent Japanese/Chinese word for dialect is 方言, which literally means "regional language/tongue". I don't think it is not at all problematic to call Ryukyuan as a regional language/tongue/word of Japan.

The real question is whether one apply the definition consistently. For example, if one assert that Ryukyuan is a language because of its unintelligibility, yet deny that Tsugaru speech is a language, then that person is being inconsistent and biased. Also, if one assert that Ryukyuan is just a subgroup of standard Japanese, that person is showing nationalist bias. As long as one recognised that there are different degree of regiona variation in japan and that difference between Ryukyu and mainland is the widest, then how you name each subcategory should be a matter of convention.

If one follow the "convention" of unintelligibility I'm fine with Ryukyuan being called language in European sense as long as major mainland dialects are recategorised as language. But given that, in Japan, hogen usually refers to unintelligible regional variation, it is an annoying hassule to call ryukyuan as genngo instead of hogen just so we look politically correct in the eyes of Western critics.

  • 3
    I think you make some good additions to the conversation, but you might want to avoid implying that the original questioner is prejudiced. Definitely, the Japanese terms can't be translated one-to-one to the English terms. But in the specific example given, The Languages of Japan by Masayoshi Shibatani, is the use of the "dialect" concept really just a translation from Japanese "hogen"? I don't know if the book was originally published in Japanese or English. Commented Jun 6, 2015 at 19:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.