Japanese authors write for Japanese readers, and not surprisingly, English authors write for English-speaking readers. Authors of all sorts use the slang, the cultural references and the rich vocabulary of their native tongues. Quite a bit of what makes a piece of literature so interesting, however, is the unspoken cultural milieu behind it, and the translator can be faced with a real problem in how to render it in another language... or whether to render it at all.

Consider the rich puns in Spider Robinson's Callahan's Crosstime Saloon series, ee cummings' use of lower case letters, or the invented jargon in William Gibson's Neuromancer.

The problem is not limited to English authors, of course. Japanese authors can't do all of the same things, because (for example) there are no upper case and lower case letters in Japanese, but they can do the same sort of thing. And they can do more, thanks to the complex writing system the Japanese language uses.

In English, emphasis can be added by using italics, bolding, or a larger font. A different font alone can make quite a difference. In most English-language literature, italics are by far the most common, possibly because having too many font gimmicks on a page makes it hard to read.

In Japanese, though, you have three or four character sets to choose from: kanji, hiragana, katakana and Roman letters. Most Japanese is written in a combination of kanji and hiragana, with katakana generally reserved for imported words that have not yet been completely naturalized into the language. There is a trend in Japan now to adopt massive numbers of foreign words, especially from English, and as a result katakana is making up an increasingly large percentage of the characters used in newspapers and books.

Katakana is also used to indicate emphasis, however, much like italics in English. If used in an unexpected fashion it can make a Japanese word look bizarre and almost foreign... Writing 愛している (ai shite iru; I love you) is the accepted way to write it, but if it is written in a combination of kanji and katakana (愛シテイル) it feels as if a robot is speaking. Needless to say, this is pretty useful because it allows the author to convey information without having to use words to explain it. If the entire sentence is written in katakana (アイシテイル) it becomes almost impossible to read, even for native Japanese speakers, because without the kanji "hook" it is difficult to tell where to start parsing the word, or even how many words there are.

Kanji can also be used for their sound rather than for their meaning, or for both simultaneously, making it possible to pack two entirely different meanings into a single kanji or kango (漢語; pair of kanji that form a unique word). This is usually handled with furigana, where the pronunciation of a kanji or kango is given above it in smaller hiragana characters. In common usage, this only happens with highly unusual kanji the reader wouldn't be expected to know, names to be sure the reader knows the correct reading, or when the material is aimed at younger readers. If the furigana is different from the dictionary pronunciation, the author has just added more information. For example, it would be possible to write 妻 (tsuma; my wife) and add furigana above it reading マイワーフ (my wife). The author has added a katakana phrase to a Japanese kanji, and all of a sudden we get the feeling that the speaker is a younger Japanese, or someone at the forefront of youth culture, or somehow involved with languages and people other than Japanese. It would be unlikely (but certainly not impossible) for, say, a 60-year old man to do this.

Another possibility is something like 暗殺心 (Assassin) by 都筑 道夫 (Tsuzuki Michio).

His book deals with kanji-based wordplay and is probably impossible to translate adequately... rather like having to explain the joke before anyone "gets" it. For example, a telepathic killer thinks in kanji, and the person receiving the message sees the kanji rather than hearing the sender's thoughts, as is the usual approach in English. But Tsuzuki has the killer use incorrect kanji, chosen both for their sound and (entirely separate) interesting meanings: The sounds, added, together, spell out "Die!" but the kanji themselves, read literally, say 紫根 (pronounced shi-ne), which would translate as "purple root." The plot in a number of the stories in this connected-story collection hinge on discovering the correct meaning from among the multiple possible ones.

But how would you translate such a thing into English?

That's a real question, because I don't have an answer. It is possible to work both meanings into the English once in a while. The former case could possibly be translated as "My tsuma, my wife," deliberately leaving in the Japanese in the same way that the original uses both languages, but of course an English-speaking reader gets a different impression.

As to the latter, Tsuzuki's wordplay... I've been searching for a good approach for about twenty years now.