Persona 5, untranslatable jokes, and tokin on 4/20
To preface this: Persona 5 is the best JRPG in years. Please buy the game. Please do not use this or any bitching about the localization you see as a reason not to get it, because it isn’t one. It’s great. Play it. I will probably be making a post soon about some of the brilliant game design things in it!
So I was irritated about one of the questions in Persona 5 and made a random tweet about it after looking it up in english sources and trying to figure out a factual basis for it. All the english sources on tokin (the name of a promoted pawn in Shogi) did not mention the history of how kin became to, and, more importantly, I could not find a single example of a cursive kin anywhere that looked anything like a to. So what the game was saying seemed pretty darn suspect.
Now to be fair, I don’t consider myself an authority on Japanese language and wouldn’t expect to be cited as such, but shortly thereafter Chris Kohler at Kotaku linked to the tweets and made an article based on them:
http://kotaku.com/this-might-be-persona-5s-biggest-translation-fail-1794223069
Now I want to specify that I was not contacted ahead of time about this. I didn’t plan or want this to become an article at a major game site. I was just kinda bitching. I did not expect this to extend beyond the people who follow me, many of whom give no shits about Persona, localization accuracy, or the minutiae of Kanji. If I had been asked, I would have preferred not to be cited for two reasons:
It’s really a minor issue and I don’t think it’s representative of my overall feelings on the game.
The level of research I put into it the tweet was appropriate for a tweet, not for an article in a major game news site.
But, at the same time, I don’t think Chris had any bad intentions with the article - it was just an interesting point of the language and an odd thing to show up in an English localized game. Chris is a good dude.
Anyways, as soon as I saw it had been made into an article, I panicked because if I wanted to be more sure I was accurate, I should have looked at Japanese sources. So I did. And, thankfully, japanese google provided this article which was massively helpful:
http://www.tonan.jp/moji/10tokin/
So, this is a long article. But what it boils down to, is arguing that it is not, in fact, a hiragana と at all - but, it also argues that it is quite uncertain as to what it actually is.
It looks like he ends up feeling like the most likely explanation is that it originated from 今, which is an ateji (phonetic replacement or spelling of a kanji that is read a certain way with no regard to its meaning) for 金.
http://tonan.seesaa.net/article/31091021.html
Anyways, either way, my explanation was wrong, which I then brought up to Chris, who was nice enough to update the article.
Now, while admitting that I was inaccurate in my factual reasoning as to why it was factually wrong, the fact remains that the statement in the english version of the game - that と金 is the cursive form of a specific kanji - is wrong. Now, it appears that there’s argument as to whether the Japanese was even correct - but it’s certainly different, and less wrong than the English. Let me explain why.
Here’s the original Japanese:
In text:
この字はある漢字の崩し字なんだ、 なにか分かる？
So, let’s break this down. First, I’ll translate the easy nouns. 字 (ji) is character (it can mean other stuff but that’s most likely in this context). 漢字 (kanji) is… kanji.
The other noun is 崩し字 (kuzushiji) , which can sort of theoretically be translated to “cursive” or “simplified” character - but it doesn’t mean cursive in really the same sense that western cursive is used. There’s a ton of possible ways to write a character in 崩し字 - really, anything that is written using a brush and omits or simplifies strokes is technically 崩し字. Which is why, in the articles above, he dislikes the argument that tokin originates from a 崩し字 of 金, because while it could be a theoretically VERY cavalier 崩し字, it really drastically simplifies the kanji (far moreso than the other shoji pieces simplify the same kanji). But I’ll give that that’s a lot of detail the game can’t provide so I’ll let it get away with cursive form :)
Okay, so we have:
この character はある kanji の cursive form なんだ、なにか分かる？
Next, the other easy stuff. この is this. なんだ is “what is”. なにか means something. So we get:
This character はある kanji の cursive form what is、something 分かる?
Last we’ll do verbs and particles. I won’t explain particles because they’re a whole section of Japanese grammar that’s not really pertinent and is somewhat complex but suffice it to say they are short “grammatical helpers” that mark the grammatical function of words in sentences, sort of like helping verbs and articles in English. Anyways, the verbs here are ある (aru), which means to be/exist (specially referring to inanimate objects), and 分かる (wakaru) which means a lot of different things in english, but usually it means do you understand or know. The only other things are は, which is a particle marking a new topic of conversation in a sentence, and has no real translatable direct equivalent in english, and の, which indicates that a noun possesses or modifies another noun. So, without any word rearrangement, you end up with:
“This character is cursive form of kanji what is, something understand?”
Which is a little intentionally obtuse but some simple rearrangement gets you something closer to english:
“This character is cursive form of what kanji, something understand?”
This is all you can really infer from transliteration. To turn this into actual English we need to start making value judgments and doing actual translation. So to get English out of this, the first and easiest thing to do is to realize that なにか分かる is just ellipsing the subject, which is probably the student who the teacher is asking the question to. Also, you need your helping verbs and pronouns, and the nanika wakaru is more of an expression, so it’s probably closer to “do you know?”. All of that is interpretation and at the translator’s discretion, but it’s necessary to end up with something that’s vaguely English.
For the first part of the sentence, you are almost actually at a readable fragment, you just need an article, and here’s where I think the translation goes wrong. It uses “the”, when it clearly should use “a”.
Why? “the” implies singularity and absoluteness. “the” implies that there is a single, or at the very least primary answer to the question “what is the cursive form of this kanji”, and that that single answer is the one they are looking for. But any Japanese person will tell you that 1) there are many cursive forms of Kanji and 2) that the “primary” cursive form of the 金, if it existed, would not be what is seen on the shogi piece. In fact, the very next slide conflicts with the usage of “the”, because it shows three different purported cursive forms of 金- and again, whether tokin factually is a cursive form of 金 is debatable, but that is at least asserted in the Japanese text, too!
So, I think a closer translation (and closer to the original Japanese structure) would be something like: “This character is a cursive form of a kanji - do you know which?” If you wanted to make it flow better, you could go from there. Something like “Can you tell me which cursive kanji this character originated from?”
Now, you might argue a simple swapping of “a” for “the” isn’t a big deal. And you would be right to some extent. But I certainly would argue that it undermines the coolest part of these little quiz games, which is that they actually teach you things! You could imagine someone playing this game, then later picking up Japanese, seeing a hiragana “to”, and telling their teacher “oh I learned that’s a cursive “kin!”” to be greeted by a frowning head-nod no. One of the coolest parts of the persona games is the opportunity to learn about Japanese, and it’s a bummer that it gives you misinformation by way of a relatively simple error.
But - I actually think the bigger translation issues, and the reason why I think a bigger change should have been made to this question, is in the answers. So, this is actually even a tricky question in japanese! While searching for japanese sources, I actually found this blog:
http://karigezima.com/archives/25714
which says:
将棋経験者ならすぐに分かる問題です。
将棋してない人からしたら難しい…かな？(´･ω･`;)
Or - “Shogi players will instantly know the answer. But people who don’t play shogi probably screamed about how hard it is, huh?”
So yeah - this is a tricky question in Japanese. But, in typical multiple choice test fashion, it actually has something to help you out:
So here’s the answer options in Japanese. Let’s compare them with the english options:
So, to start things off, the question is asking about the cursive form of kanji - which you need to be able to see the kanji to determine, but the english version doesn’t let you see the freaking kanji! And, a single English word might correspond to multiple kanji. For instance, divination as a kanji could just as easily be 占 as卜. And, while it’s more of a technicality because 金 is certainly the most obvious kanji that means “gold”, gold could also mean 鋆, 釛 , 鉚, 鎏 , 鏐 . Those might not be things that a sane person would think of when given the option “gold”, the fact remains that the game is supposed to be translated into English and you shouldn’t need to know Kanji to understand and answer questions correctly in it!
But I think the bigger deal here is that you actually miss out on the “dummy” option. See, because while “Divination” is technically what 卜 means, it is not an option in the Japanese version because of that - it’s there because it’s looks almost exactly like ト, which is the same sound, “to”, in the other major japanese character system, Katakana.
So for a Japanese speaker, you instantly see that dumb middle option and rule it out because it’s clearly a trick answer for dummies. It’s even maybe a little chuckle-worthy. You just miss out on all of that in english.
Because it’s so untranslatable to someone who doesn’t understand Japanese, and because really the relevant point here narratively and symbolically is that the lowest shogi piece can promote into the Gold General, and how that ties into the idea of neauvau riche, I think it would have been nice to just ask something like “This piece indicates when a pawn is promoted to what rank”? It’s a bit further from the Japanese, but again, if you can’t convey the meaning in Japanese without explaining what kanji are, how they are simplified, and make the answers make sense visually to players, maybe just get at the heart of the issue and the bigger narrative point.
Anyways, long story short - I still think this could have been better, but I’m glad it existed in a way because it was very much a learning opportunity for me!
After posting this, someone added a helpful comment:
Hey, Nate. MDB again. Thought I could lend my expertise, since it’s always better to have two experts tackling a problem like this. Here are some notes:
なんだ (nanda) does not mean “what is.” It’s the short/informal form of the phrase なのです (nano desu), which is simply a way to end a clause that ends in a noun while also inserting your personal feelings into the phrase. In effect, it’s a verbal punctuation mark. If that sounds terribly confusing, that’s because it’s Japanese, one of the most complicated languages on Earth.
The word ある (aru) is, as you said, a verb meaning “to be/exist.” And it’s true that placing a verb before a noun makes it a relative clause. So you technically could say that the translation for ある漢字 is “the kanji that exists.” But that extremely inefficient and redundant. You’d get the exact same meaning from taking the “aru” out, so there’s no reason for it to be there.
Which is why! The word “aru” before a noun is used to indicate a particular one of those things. So ある日 means “someday,” ある人 means “somebody,” etc. So the proper translation for ある漢字 would be “some (particular) kanji.”
なにか (nanika) does indeed mean “something” or “anything,” but in the particle か (ka) indicates a request for specific information. When asking a question with a yes or no answer, you use かどうか (kadouka). Examples:
「いくらか分かる？」(Ikura ka wakaru?) “Do you know how much (it is)?”
「どこにあるか全く知らんねんで。」(Doko ni aru ka mattaku wakannen de.) “I ain’t got a clue where it is.”
なに (nani), as I’m sure you know, means “what,” so when the teacher asks, “Nani ka wakaru?” she’s not saying, “Something understand?” it’s more like, “Do you know what (it is)?”
Finally, in your post, you accidentally omitted part of the Japanese version. You wrote:
「 この字はある漢字の崩し字なんだ 、なにか分かる？」
But the text in the game actually says:
「 この字はある漢字の崩し字なんだけど、なにか分かる？」
It’s not super important, but it still holds a minor significance that has to be considered depending on how you want to convey the character. Personally, I like to translate けど (kedo) to “Even though”, but the word “but” works just as well. It’s much more complicated than that, so just role with it, for now. Get it? “Role”? ‘Cause it’s an RPG? Um…anyway.
When you put it all together, there are several ways you could translate the sentence, but it turns out the way it was localized was actually pretty faithful: “This character is the cursive form of a specific kanji. Do you know which one it is?” A little verbose compared to the original, but grammatically, very close!
Hope that helps!
Thanks! That’s helpful :D It’s good to talk about this stuff.
You’re right about なんだ in this context, for sure. I can be a casual form of なんです, which is what had confused me.
The other points are totally valid, and I totally agree I missed a couple things. I still think the choice of “the” is wrong, though as I said I think it’s a pretty minor issue. I still wish they showed you the kanji options though!
It’s really a minor issue and I don’t think it’s representative of my overall feelings on the game.
The level of research I put into it the tweet was appropriate for a tweet, not for an article in a major game news site.
But, at the same time, I don’t think Chris had any bad intentions with the article - it was just an interesting point of the language and an odd thing to show up in an English localized game. Chris is a good dude.
Anyways, as soon as I saw it had been made into an article, I panicked because if I wanted to be more sure I was accurate, I should have looked at Japanese sources. So I did. And, thankfully, japanese google provided this article which was massively helpful:
http://www.tonan.jp/moji/10tokin/
So, this is a long article. But what it boils down to, is arguing that it is not, in fact, a hiragana と at all - but, it also argues that it is quite uncertain as to what it actually is.
It looks like he ends up feeling like the most likely explanation is that it originated from 今, which is an ateji (phonetic replacement or spelling of a kanji that is read a certain way with no regard to its meaning) for 金.
http://tonan.seesaa.net/article/31091021.html
Anyways, either way, my explanation was wrong, which I then brought up to Chris, who was nice enough to update the article.
Now, while admitting that I was inaccurate in my factual reasoning as to why it was factually wrong, the fact remains that the statement in the english version of the game - that と金 is the cursive form of a specific kanji - is wrong. Now, it appears that there’s argument as to whether the Japanese was even correct - but it’s certainly different, and less wrong than the English. Let me explain why.
Here’s the original Japanese:
In text:
この字はある漢字の崩し字なんだ、 なにか分かる？
So, let’s break this down. First, I’ll translate the easy nouns. 字 (ji) is character (it can mean other stuff but that’s most likely in this context). 漢字 (kanji) is… kanji.
The other noun is 崩し字 (kuzushiji) , which can sort of theoretically be translated to “cursive” or “simplified” character - but it doesn’t mean cursive in really the same sense that western cursive is used. There’s a ton of possible ways to write a character in 崩し字 - really, anything that is written using a brush and omits or simplifies strokes is technically 崩し字. Which is why, in the articles above, he dislikes the argument that tokin originates from a 崩し字 of 金, because while it could be a theoretically VERY cavalier 崩し字, it really drastically simplifies the kanji (far moreso than the other shoji pieces simplify the same kanji). But I’ll give that that’s a lot of detail the game can’t provide so I’ll let it get away with cursive form :)
Okay, so we have:
この character はある kanji の cursive form なんだ、なにか分かる？
Next, the other easy stuff. この is this. なんだ is “what is”. なにか means something. So we get:
This character はある kanji の cursive form what is、something 分かる?
Last we’ll do verbs and particles. I won’t explain particles because they’re a whole section of Japanese grammar that’s not really pertinent and is somewhat complex but suffice it to say they are short “grammatical helpers” that mark the grammatical function of words in sentences, sort of like helping verbs and articles in English. Anyways, the verbs here are ある (aru), which means to be/exist (specially referring to inanimate objects), and 分かる (wakaru) which means a lot of different things in english, but usually it means do you understand or know. The only other things are は, which is a particle marking a new topic of conversation in a sentence, and has no real translatable direct equivalent in english, and の, which indicates that a noun possesses or modifies another noun. So, without any word rearrangement, you end up with:
“This character is cursive form of kanji what is, something understand?”
Which is a little intentionally obtuse but some simple rearrangement gets you something closer to english:
“This character is cursive form of what kanji, something understand?”
This is all you can really infer from transliteration. To turn this into actual English we need to start making value judgments and doing actual translation. So to get English out of this, the first and easiest thing to do is to realize that なにか分かる is just ellipsing the subject, which is probably the student who the teacher is asking the question to. Also, you need your helping verbs and pronouns, and the nanika wakaru is more of an expression, so it’s probably closer to “do you know?”. All of that is interpretation and at the translator’s discretion, but it’s necessary to end up with something that’s vaguely English.
For the first part of the sentence, you are almost actually at a readable fragment, you just need an article, and here’s where I think the translation goes wrong. It uses “the”, when it clearly should use “a”.
Why? “the” implies singularity and absoluteness. “the” implies that there is a single, or at the very least primary answer to the question “what is the cursive form of this kanji”, and that that single answer is the one they are looking for. But any Japanese person will tell you that 1) there are many cursive forms of Kanji and 2) that the “primary” cursive form of the 金, if it existed, would not be what is seen on the shogi piece. In fact, the very next slide conflicts with the usage of “the”, because it shows three different purported cursive forms of 金- and again, whether tokin factually is a cursive form of 金 is debatable, but that is at least asserted in the Japanese text, too!
So, I think a closer translation (and closer to the original Japanese structure) would be something like: “This character is a cursive form of a kanji - do you know which?” If you wanted to make it flow better, you could go from there. Something like “Can you tell me which cursive kanji this character originated from?”
Now, you might argue a simple swapping of “a” for “the” isn’t a big deal. And you would be right to some extent. But I certainly would argue that it undermines the coolest part of these little quiz games, which is that they actually teach you things! You could imagine someone playing this game, then later picking up Japanese, seeing a hiragana “to”, and telling their teacher “oh I learned that’s a cursive “kin!”” to be greeted by a frowning head-nod no. One of the coolest parts of the persona games is the opportunity to learn about Japanese, and it’s a bummer that it gives you misinformation by way of a relatively simple error.
But - I actually think the bigger translation issues, and the reason why I think a bigger change should have been made to this question, is in the answers. So, this is actually even a tricky question in japanese! While searching for japanese sources, I actually found this blog:
http://karigezima.com/archives/25714
which says:
将棋経験者ならすぐに分かる問題です。
将棋してない人からしたら難しい…かな？(´･ω･`;)
Or - “Shogi players will instantly know the answer. But people who don’t play shogi probably screamed about how hard it is, huh?”
So yeah - this is a tricky question in Japanese. But, in typical multiple choice test fashion, it actually has something to help you out:
So here’s the answer options in Japanese. Let’s compare them with the english options:
So, to start things off, the question is asking about the cursive form of kanji - which you need to be able to see the kanji to determine, but the english version doesn’t let you see the freaking kanji! And, a single English word might correspond to multiple kanji. For instance, divination as a kanji could just as easily be 占 as 卜. And, while it’s more of a technicality because 金 is certainly the most obvious kanji that means “gold”, gold could also mean 鋆, 釛 , 鉚, 鎏 , 鏐 . Those might not be things that a sane person would think of when given the option “gold”, the fact remains that the game is supposed to be translated into English and you shouldn’t need to know Kanji to understand and answer questions correctly in it!
But I think the bigger deal here is that you actually miss out on the “dummy” option. See, because while “Divination” is technically what 卜 means, it is not an option in the Japanese version because of that - it’s there because it’s looks almost exactly like ト, which is the same sound, “to”, in the other major japanese character system, Katakana.
So for a Japanese speaker, you instantly see that dumb middle option and rule it out because it’s clearly a trick answer for dummies. It’s even maybe a little chuckle-worthy. You just miss out on all of that in english.
Because it’s so untranslatable to someone who doesn’t understand Japanese, and because really the relevant point here narratively and symbolically is that the lowest shogi piece can promote into the Gold General, and how that ties into the idea of neauvau riche, I think it would have been nice to just ask something like “This piece indicates when a pawn is promoted to what rank”? It’s a bit further from the Japanese, but again, if you can’t convey the meaning in Japanese without explaining what kanji are, how they are simplified, and make the answers make sense visually to players, maybe just get at the heart of the issue and the bigger narrative point.
Anyways, long story short - I still think this could have been better, but I’m glad it existed in a way because it was very much a learning opportunity for me!
After posting this, someone added a helpful comment:
Hey, Nate. MDB again. Thought I could lend my expertise, since it’s always better to have two experts tackling a problem like this. Here are some notes:
なんだ (nanda) does not mean “what is.” It’s the short/informal form of the phrase なのです (nano desu), which is simply a way to end a clause that ends in a noun while also inserting your personal feelings into the phrase. In effect, it’s a verbal punctuation mark. If that sounds terribly confusing, that’s because it’s Japanese, one of the most complicated languages on Earth.
The word ある (aru) is, as you said, a verb meaning “to be/exist.” And it’s true that placing a verb before a noun makes it a relative clause. So you technically could say that the translation for ある漢字 is “the kanji that exists.” But that extremely inefficient and redundant. You’d get the exact same meaning from taking the “aru” out, so there’s no reason for it to be there.
Which is why! The word “aru” before a noun is used to indicate a particular one of those things. So ある日 means “someday,” ある人 means “somebody,” etc. So the proper translation for ある漢字 would be “some (particular) kanji.”
なにか (nanika) does indeed mean “something” or “anything,” but in the particle か (ka) indicates a request for specific information. When asking a question with a yes or no answer, you use かどうか (kadouka). Examples:
「いくらか分かる？」(Ikura ka wakaru?) “Do you know how much (it is)?”
「どこにあるか全く知らんねんで。」(Doko ni aru ka mattaku wakannen de.) “I ain’t got a clue where it is.”
なに (nani), as I’m sure you know, means “what,” so when the teacher asks, “Nani ka wakaru?” she’s not saying, “Something understand?” it’s more like, “Do you know what (it is)?”
Finally, in your post, you accidentally omitted part of the Japanese version. You wrote:
「 この字はある漢字の崩し字なんだ 、なにか分かる？」
But the text in the game actually says:
「 この字はある漢字の崩し字なんだけど、なにか分かる？」
It’s not super important, but it still holds a minor significance that has to be considered depending on how you want to convey the character. Personally, I like to translate けど (kedo) to “Even though”, but the word “but” works just as well. It’s much more complicated than that, so just role with it, for now. Get it? “Role”? ‘Cause it’s an RPG? Um…anyway.
When you put it all together, there are several ways you could translate the sentence, but it turns out the way it was localized was actually pretty faithful: “This character is the cursive form of a specific kanji. Do you know which one it is?” A little verbose compared to the original, but grammatically, very close!
Hope that helps!
Thanks! That’s helpful :D It’s good to talk about this stuff.
You’re right about なんだ in this context, for sure. I can be a casual form of なんです, which is what had confused me.
The other points are totally valid, and I totally agree I missed a couple things. I still think the choice of “the” is wrong, though as I said I think it’s a pretty minor issue. I still wish they showed you the kanji options though!