Victor Mair has a Log post about an interesting situation:
“Arirang” (Hangul: 아리랑) is arguably the most famous Korean folk song. Indeed, “Arirang” is so well-known that it is often considered to be Korea’s unofficial national anthem. Yet no one is sure when the song arose nor what the title means. […]
There are hundreds of theories of the origin and meaning of “arirang”. In “What Does Arirang Mean? The Theories on the Etymology of Arirang” (5/24/15), the author examines nine of the theories, which ascribe the song’s origin to dates ranging from the first c. BC to the late nineteenth century AD and which contend that the title is based on the personal name of two different heroines, that it means “I Part from My Dear”, that it means “Our Escape Is Difficult”, that it means “My Ears Become Deaf”, that it means “Mute and Deaf”, that it is a Classical Chinese onomatopoeic expression signifying the grunts of laborers, that it signifies “Russia, America, Japan, and England” (!), or that it is the name of a hill. The phonological transformations that are required to get from many of these terms and expressions to “arirang”, quite frankly, require considerable imagination.
The linked post by “Kuiwon” sounds somewhat tendentious, and frankly I’m happy to leave the song’s origin a mystery, but the discussion is worth reading, and I was particularly struck by this comment from Bob Ramsey:
[…] After all, Korean etymological “science” itself is pretty free-wheeling—just as the corresponding studies in Japan are. Virtually all of the so-called etymological dictionaries you see in Korea are ridiculously fanciful and often outrageously anachronistic. And that is certainly true of everything you see said about the word arirang. I doubt it’s a mystery that will ever be solved.
I might point out, though, that my old mentor Lee Ki-Moon has for decades now been engaged in a truly serious project of putting together a genuine etymological dictionary of Korean But his work has been excruciatingly slow and difficult. The etymologies that he’s written up and sent me are always carefully documented, but for the most part they are studies about obscure and sometimes obsolescent words. I’m thinking that his project will outlast him and never be completed.
What a sad situation! I realize I’m spoiled being a speaker of a language with over a century of scientific etymological work and a student of other such languages, but it always shocks me to learn about major languages (Arabic being, I suppose, the most prominent) with no good etymological resources. Korean has (according to Wikipedia) about 80 million speakers, for heaven’s sake; it should have at least one decent etymological dictionary, even if it only covered basic vocabulary.
To be fair, English has well-known relatives and borrowing-source languages. For Korean, we don’t even know if it’s an isolate or not, and the only clear source of borrowing before modern times is Chinese. So for most native words there’s probably little or nothing to say.
Nothing excuses the neglect of Arabic, to be sure.
Well, you could certainly give the earliest known or reconstructable forms, and of course the Chinese and Japanese originals of borrowed words.
I can confirm that Japanese etymology has been in a pretty sorry state, but the new Nikkoku dictionary makes a decent stab in the right direction. You can see in the pink highlighted part of the sample below that after listing a few possible origins of “akatoki” (archaic “morning”) they even posit an original pronunciation: “Ak-Tok”!
http://www.nikkoku.net/introduction/mihon_09.html
You could show older Korean forms, yes, but that would be more a historical dictionary than an etymological one. (Not that that’s a bad thing.) For Japanese, there wouldn’t be much point in saying that the etymology of (kanji) is (hanzi), any more than the Dictionary of the Scots language has entries like house ‘house’.
I disagree to an extent with you, John, there would actually be quite a bit of value in a Japanese etymological dictionary with entries written in a way (presumably Roman characters) that made sound changes clear. The thousand-year-reign of kana is a blessing and a curse: on the one hand, it makes ancient texts easier to read, but on the other, it hides how much the language has changed, suggesting for example that “house” has been “ie” since the beginning of time when it is very well known that it was much closer to “ipye”. (Of course, there would be resistance to trying to revive older pronunciations for older texts, since there is such a long tradition of quoting them within newer texts, similar to Chinese and even English to an extent — “historically informed pronunciation” of Shakespeare is still very much a fringe thing.)
Marc: The NKD is a great resource and their “etymological theories” (語源説) section is always an interesting read, but the problem is that they don’t offer much guidance on how to interpret it. It’s usually just a list of all the vaguely plausible theories they can find. The “Ak-Tok” pronunciation for example is credited to Yosano Hiroshi, better known as Yosano Tekkan, even better known as “Yosano Akiko’s husband” — even by the standards of the time, he was more journalist and poet than linguist, and he published his book in 1931. That doesn’t mean he was wrong, of course, but the inclusion of the reconstruction makes his theory look the most “scientific” of the bunch when really this is not the case. I think that dictionary users would be better served by some editorial commentary: “the current academic consensus favors X” or “is split between X and Y” or “no consensus has been reached, but it is agreed that Z is unlikely.”
Of course because Japanese is in a similar position to Korean, probably for the vast majority of cases the note would just say “We have no idea, it could be any of these.” Even that would clarify things, though.
Things are getting better for Japanese etymology though. In particular interest in the Ryukyuan languages is growing (along with recognition that by any reasonable definition they are separate languages, not dialects of Japanese), and eventually that should lead to better (or at least more easily testable) reconstructions of proto-Japonic. There seems to have been a lot of interest in Old Japanese recently, including (especially?) in the English-speaking world, and that might help too. And of course it’s always possible that Alex Ratté’s forthcoming work (see recent Altaica thread) will show some convincing correspondences between Japanese and Korean, or some earlier Korean-peninsula language (I’m not sure exactly what Ratté’s theory is) — that would obviously be revolutionary for both Korean and Japanese etymology!
There are a number of “historical” dictionaries as John Cowan calls it, which gives the historical forms of words. Even the Pyojun Gugeo Daesajeon 표준국어대사전 “Great Dictionary of Standard Korean” which is a general dictionary of the language includes the earliest attested forms for lots of words, and I find it a valuable resource. It also includes reconstructions of unattested forms and etymologies for certain entries. A couple of samples:
For 붓 but “brush”, it has 【＜붇＜훈해＞＜筆】. This means that 붓 comes from the form 붇 but attested in 훈민정음 해례본 Hunminjeongeum Haeryebon (1446), which itself in turn derives from the Chinese () 筆.
For 보라매 boramae “a young hawk tamed for hawking”, it has【보라매＜훈몽＞←보라 [＜boro] +매】. This means that the earliest attested form is already 보라매 in the 훈몽자회 Hunmongjahoe (1527), and that the etymology is 보라 bora + 매 mae where 보라 comes from Mongolian () boro.
Out of necessity, the earliest historical forms only go up to the first publications in hangul, the Korean alphabet, in the 15th century. Before hangul, Korean could only be represented with Chinese characters in complex systems such as 향찰 hyangchal and 이두 idu, which are difficult to interpret because it is not always clear when a particular Chinese character is being used for its sound or meaning, and they don’t give much phonetic information anyway.
This state of affair is also a consequence of the fact that Korean is a language isolate; once you have taken out the obvious Chinese loanwords, the Khitan+Mongolian+Manchu loanwords, the older Chinese loanwords, the problematic and very small common Japanese/Korean vocabulary (which contains loanwords from both sides, as well as a residue that might be indicative of a very remote genetic relationship), and the productive language-internal derivations, not much remains to be done, especially for Korean (at least, in Japanese you have the Ryûkyûan languages). Without related languages, reconstructing non-productive derivational morphology has severe limitations.
the fact that Korean is a language isolate
The theory, rather: there is another theory, which makes it hard to do an etymological dictionary with the necessary single-mindedness to see it through.
Hi everyone,
I know I’m a bit late to the party here, but I just discovered this blog tonight and was very excited to come across a discussion of Korean etymology. I’ve been working as a Korean translator for a year and a half, and have always been interested in etymology in general. This year, I found a Korean etymological dictionary which seems quite well-written, but then I’m not a professional etymologist or historical linguistic so I’ll leave it up to you to decide. Here is the author and title: 백문식 “우리말 어원 사전” (published in 2014). As I said, I’m not a professional, but so far it’s much more convincing than others such as “뜻도 모르고 자주 쓰는 우리말 어원 500 가지” which seem like collections of folk etymologies and stories (entertaining to read, no doubt, but frustrating when you want something concrete). I’d love to hear anyone’s thoughts on it.
Welcome, Mathieu — I’m glad you found the site and the post! I hope someone who actually knows something about the subject can respond about the dictionary.
Just got to this thread via Random Link, and I’m sad to see nobody responded to Mathieu — I’m still baffled that there’s so little available on Korean etymology!