[Jongseong]

I'm dismayed that this piece is repeating Marc Rogers's gross misportrayal of the linguistic situation of the Koreas, saying that "Korean language in the code also suggests a Korean origin, though not necessarily a North Korean one, since North Koreans use a unique dialect." First of all, North Korean doesn't have "a unique dialect" but a number of regional dialects, just like South Korea, and like the situation in many languages. But again as in many major languages, a supra-regional, standard Korean language came into being, based on the central dialect region around Seoul which was the capital for many centuries. Before that the capital was Kaesong, which is in the same central dialect region as Seoul though it is now in North Korea. This happened before the division of the peninsula. Even today, the standard Korean taught and spoken in North Korea is based on this common standard with the South. The differences between regional dialects within either North or South Korea are far greater than the difference between the standard Korean spoken in the North and the South. The difference is mainly in words (especially any technology-related vocabulary introduced after the end of WWII) and spelling, and it's a lot like the differences between British and American English. You're never going to say that something written in English can't have been written by Americans because they have a unique dialect.

Also, as far as I know the codes didn't contain any Korean. Instead, what they found was that it seems to have used Korean text encoding, like EUC-KR. People have pointed out that this is a South Korean encoding, but North Koreans also use it since you hardly find any software that supports the official North Korean encoding. Again, if someone uses a British English locale, that isn't proof that it can't be an American. When it comes to text encoding and locale, you usually use whatever is available that lets you type in your own language.

[Jongseong]

To understand the language situation in Korea, let's imagine that we divided the Italian peninsula in half just north of Rome. Does it follow that now there are two dialects, North Italian and South Italian? No. Italian has a bewildering variety of regional dialects, and our arbitrary line doesn't correspond to a genuine dialect border. Similarly, the DMZ cuts across the central dialect region in Korea.

More importantly, Italians will continue to write and be taught in Standard Italian, which was developed based on the Tuscan dialect long before our artificial division of the peninsula. It won't be as if they would start from scratch and create new standard languages based on the Milanese dialect in the North and the Roman dialect in the South. Even independent countries such as Germany, Austria and Switzerland find it useful to use the same standard German as each other, even if it's not necessarily based on a dialect spoken wothin their borders. There will inevitably be differences in vocabulary and spelling, but the differences will be far less than if we imagined a naïve model where each country creates its own standard (which is what basically happened in Scandinavia).

[blahblah7777]

Your comparison of Korean to Italian isn’t helpful, because you’re confusing dialects and languages. Italy has many regional languages including Piedmontese, Lombard, Venetian, Sicilian, Sardinian, Neapolitan, etc., and of course Italian (based on the Tuscan language). Each regional language has many dialects, so for example in Piedmont you have the Turinese dialenct of Piedmontese from Turin, which is distinct from the Piedmontese dialect in other areas. The dialects vary substantially, from one village to another—even different suburbs of a city can have different words for certain things.

Confusingly, Italians would call Piedmontese or Lombard a “dialetto” as much as they would call Turinese a “dialetto”. The word basically means a dialect or regional language, depending on the context. There is also a political element to it—the Italian government has suppressed the regional languages for years, and even now does not recognise them as languages, against academic opinion.

To be clear, Italians would also (generally) refer to Welsh as a “dialetto” of English, despite the fundamental difference of Welsh and English. (In fact they would usually also often refer to the U.K. as “inghilterra”.) The word “dialetto” as currently used in normal Italian speech simply does not correspond 100% with the English word “dialect”, much like the word “camello” doesn’t correspond to “camel”.

The regional languages generally are not mutually intelligible, although this depends on which dialects two speakers speak, and how “stretto” (strong) the dialect is (I don’t know what the academic term for this is). So for example Vercellese (from Vercelli) is linguistically close to Novarese (from Novara) even though Vercellese is classed as Piedmontese and Novarese is classed as Lombard (despite being a Piedmontese city). The distinction is ultimately arbitrary—there is a gradation of dialects from Piedmontese to Lombard. Vercellese for example has many grammatical elements of Lombard (e.g. it uses the Lombard lü (meaning “he” or “him”) instead of the Piedmontese chiela).

Also, an older or more rustic speaker is more likely to speak a “stretto” dialect, because they’ll use more words and expressions originally belonging to that dialect (or to the regional language). Over the years, the regional languages have absorbed many words from Italian, replacing the traditional words. Now, the same thing is happening to Italian with English words (e.g. the word “goal” replacing “rete”, or “babysitter” replacing “tata”, or “shopping” replacing “spesa”—the English word in each case sounds more modern or cool to Italian speakers).

The linguistic situation is basically the same as with Catalan and Spanish. Catalan is as much a “dialect” of Spanish as Piedmontese would be a “dialect” of Italian. In fact, you could just as rightfully say that Italian is a “dialect” of Piedmontese. The difference is political, not academic.

[ZanyProgrammer]

From what I remember, the main differences between the DPRK and ROK in terms of language are that the North doesn't use Sino-Korean words ("pure" Korean tends to predominate), and the vast majority of post-war loan words obviously aren't used.

[Jongseong]

You've got the gist, but North Korea does use tonnes of Sino-Korean words, which are like words of Greek and Latin origin in English. You can't simply eliminate such words from the language altogether. What you can do is to favour pure Korean words over Sino-Korean when coining new terms or standardizing terms for technical usage. So North Korea might use a bit less Sino-Korean than the South (I'm not even sure though, because there are other factors such as loanwords in South Korea displacing Sino-Korean words there). But this makes barely a blip on the language as a whole since so many basic words in Korean are Sino-Korean. Even most North Korean terms you're ever likely to hear are Sino-Korean: Juche, Son'gun, Chollima, Rodong missiles...

[disputin]

" if someone uses a British English locale, that isn't proof that it can't be an American" -- I'd take that further and ask, if this was such an advanced attack, wouldn't a bit of language misdirection be thrown in?

[samweinberg]

On a somewhat unrelated note, some people believe this is what the founder of Bitcoin did.

> Both his forum posts and his comments in the bitcoin source code used such Brit spellings as optimise and colour.[1]

1: http://www.wired.com/2011/11/mf_bitcoin/all/

[feraloink]

Sigh, sounds very familiar, like what people say about Mandarin and Cantonese, those who don't know.

This is a different hack, but I think they more reliably differentiated between North Korea and South Korea, due to the IP addresses? "Korea seeks U.S. help in reactor hacking probe" http://m.koreaherald.com/view.php?ud=20141222001202&ntn=0

[Jongseong]

From what I've seen in the South Korean media, no one is talking about North Korean involvement in this separate hacking incident of nuclear reactors. The codes seem to be different from those used in previous attacks blamed on North Korea.

Note that linguistic evidence doesn't come into this, just as one wouldn't usually be speculating about which English-speaking country a hacker was from simply based on the code.

[curiously]

A good writeup on explaining the language diversity in the Korean peninsula.

What's interesting about North Korea is that they force Korean words on everything. For example, a word will spell the same in South Korea and can only be differentiated by context, intonation and use of chinese characters. Newspapers in South Korea is a good example of this as it almost requires a basic knowledge of traditional Chinese characters. North Korea seems to have none of it and has a political agenda to "purify" the language and it leads to lot of weird looking Korean words and leads to confusion.

Kaesong was also the capital a long ass time ago during the Koryo dynasty which was overthrown in a coup and Chosun dynasty was created by a general (we'll see the same thing in 1960s creating the modern ROK). The longest reigning and oppressive regime. North Koreans still refer to themselves as Chosun people and much of modern Korean identity lends from this era. The word Korea also comes from Koryo or Coree in French.