Friday, June 26, 2009

Learning Japanese, Part 2a - Looking up Kanji

ACSiren raised a good question to my last entry. What to do about kanji that have different readings according to context?

A very simple example is "yama", or "mountain". 山 can be read as "yama" by itself, or "sen" or "san" when in combination with other kanji (such as 富士山 for Fuji-san, as in Mount Fuji). The question then becomes, when do we use "sen" and when do we use "san" in determining the reading for this kanji? It's easy enough to decide when to use "yama" and "san" - "yama" is "mountain", so when you're just talking about mountains, 山 will be by itself and you'll always read it as "yama". "san" is closer to "Mt." so 山 will occur along with the other kanji for the name of a mountain, and in these cases, we'll pronounce 山 as "san", in terms of "Mount Rushmore" (Rushmore山) and "Mount Fuji".

So, what do we do about "sen"? Turns out that while "sen" does show up in the dictionaries as one of the pronunciations for 山, it's used very, very rarely in real life. In general, we can say that it's just a regional variant on "san", and we have to be told by someone else when it's to be used.

Which brings us to family names. Generally, family names are very straight forward - 山本 is always read as "Yamamoto" (not Sanpon). Through common usage and experience, we eventually get "trained" to expect a specific pronunciation for each kanji when it's used for a person's name. But, even Japanese natives can get confused when faced with a rarely-used kanji or reading. In these situations, the speaker needs to tell us the answer directly. In manga, this is done by putting the hiragana reading alongside the kanji (i.e. - furigana).

What general rules can we follow to get a specific reading for a specific kanji? First, check if the kanji is isolated from other kanji while also being followed by hiragana. If it is, it's probably being used as a verb, and the hiragana will tell you the reading to use (行きません is read as "ikimasen", or "to not go", meaning that 行 is being treated as a verb here). If it is used in combination with other kanji, the reading will be forced on you because it's basically a spelling issue (although it doesn't look like it, in 1行おき書いた ("I wrote on every other line"), the 行 character is to be pronounced as "kou" here, because it's in combination with "1", a westernized version of "一行" ("1 line"), which is either read as "ikkou" or "ichigyou" depending on which you prefer. And here, if the author wanted you to use "ichigyou", they'd put the furigana next to the kanji; ikkou is probably the more commonly-used pronunciation.)

Ok, all of the above suggestions revolve around two assumptions: first, that you're trying to figure out the readings while looking at the text itself; second, that you're not using the tools that I described in the previous post. If you have a Casio electronic dictionary, you can try writing in an entire kanji string using the stylus, and the Casio will tell you how the kanji is pronounced along with it's meaning in that context. But, often the Casio can't figure out what you're writing, or it doesn't have an entry for the tense of the verb you're trying to look up, so it's time to go to NJStar. NJStar has a feature that lets you select the different elements of the kanji you're looking for (the radicals) and then select your kanji from the list of those that use those radicals. Once you've picked that kanji and typed in the sentence the kanji is used in, the pop-up dictionary will offer suggested readings for the entire string. While NJStar is lousy at anticipating the kanji you want as you enter it, it is good at looking up entire strings in the dictionary afterwards.

Occasionally, there is the oddball case where more than one reading (and therefore, more than one definition) can be used in a specific sentence. Fortunately, the readings do tend to vary based on the desired meaning, so if you look at the rest of the sentence in context, you can figure out what meaning the writer wanted to use, and that should also give you the reading that fits best at the same time. And, NJStar's pop-up dictionary is really helpful in this situation as well.

Now, what makes a really big difference here is the expected reading level of the document you're trying to translate. If you start out with something easy, like the Ranma 1/2 manga, most of the kanji will have furigana, telling you what the readings are supposed to be. So, if you're having trouble dealing with more complicated documents, try practicing with something simpler that uses lots of furigana and then work up from there. Keep in mind, though, that manga that uses furigana will also use very simple syntax and concepts. And the entire point of this blog series is to be able to understand how the Japanese think by analyzing more difficult sentences, and of course difficult sentences are never accompanied by furigana.

No comments:

Post a Comment