Today's lunchtime, I was talking that lots of Japanese started exchanging e-mails and browsing web sites by cellular phones when a friend of mine pointed out Japanese could easily adapt to the small screen of cellular phones because of Kanji characters. Hmm... I wonder if it's the case or not. I've seen some Japanese translations of English literatures tends to be much thicker than originals, like two to four volumes of translation for a single original paperback.

Let's see an example. I compared size of Linux HOWTO documentation files and their Japanese translations. Out of 65 files, the average of the ratio of Japanese version over English was 0.95. Considering that Japanese characters usually occupy twice of the screen space as English ones, two languages seem to be comparable.

However, Japanese can further reduce number of characters by using Kanji phrases where one or two words are more likely than a phrase or a sentence, such as menus; and it can also pack information efficiently in a limited width of column because you can wrap Japanese words at almost anywhere. Japanese is very flexible to compose a complex word by just concatenating Kanji phrases, while you need noun and adjective phrases in English. If a context is known, only one or two Kanjis are enough to construct menu items. For example, In Nikki-engine's newly updated page you can see a list of one character menus ("新", "N", "マ", "覧", "索", etc.) and a repeated visitor of the site can easily grasp what they mean. In English I guess you need at least four characters ("New", "TopN", "MyNE", "List", "Srch", etc.., maybe).