Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

"is point is that learning kanji presents two obstacles: remembering what the shapes mean and remembering how they are pronounced."

How is this different from English words? How is conscientiousness pronounced? Not to mention rendezvous.



English is written using an alphabet. Alphabets are by and large phonetic so you do not need to remember how to pronounce every word, you learn to read the alphabet and specific spelling (e.g. 'th' in English) and you are mostly done.

Chinese is written using ideograms ('kanji' in Japanese), which convey a meaning but not a pronounciation. So when you encounter a new character you cannot pronounce it.


> Chinese is written using ideograms ('kanji' in Japanese), which convey a meaning but not a pronounciation. So when you encounter a new character you cannot pronounce it.

This isn't quite true - about 80% of Chinese characters are so-called phonosemantic compounds, where people originally started using the character for one thing for another thing whose word sounded similar (say, emoji for "can", as in able to) and then adding a semantic component to differentiate the character from other similar-sounding ones. In Chinese, they smushed the two components into the space of one character, but in eg. Egypt, they simply wrote whe semantic clarifier and the phonetic hint side by side, full size.

That is, the majority of the characters are primarily sound-based, it's just that the connection between a character and its sound is shoddy, even in Chinese languages.

Japanese kun readings for native words do divorce the characters pretty completely from their sound.


This is being pedantic.

You cannot guess the pronunciation of a character from the way it is written. At most the "phonosemantic compound" might provide a clue to possibly reduce the space of possibilities.


It's not pedantic in that the characters themselves really are sound-based and provide a pretty decent clue. That said, it's still just a clue, and there are multiple similar pronunciations associated with any phonetic component, so it's still guesswork in the end.


And so we are back to square one: it is not possible to know how to pronounce a character by looking at it (and any clues are not "decent" in the majority of cases), as opposed to alphabetic systems...


The difference is that for most words you can easily get 90% there in English and, even if you don't read it properly, the listener can somewhat get what you are trying to say.

And say you don't know how to read "conscientiousness" but you know how to read "con" and "ousness" you can try to go for con-shu-tiouness and at least you're somewhere.

It's not the case with Kanjis, sure you might know one part of a word but you might be wrong. Also similar looking kanji don't read similarly at all, so even that is out of the window.

And don't get me started on the approach of applying memoization techniques purely on the strokes of a kanji, that's gonna hurt more than not.

Simple examples:

- 末 and 未

- 大 and 犬

- 千 and 干


It actually works the same way in Japanese. Radicals give rough clues about meanings as well as pronunciation. Similar looking kanji do read similarly. It's just... you can read pronounce "brochures" in either French style or in Southern American, and which one is more appropriate depends.

edit: Simple examples:

  - 臆 and 億
  - 誤 and 語
  - 諦 and 締


Not always, as I wrote in another comment this is because of the Japanese script reform for "modernising" and simplifying the script. So sometimes kanji look similar but are not. More info at: https://en.m.wikipedia.org/wiki/Japanese_script_reform#Mazeg...


So, I don't know what to say, but to be honest, I never even thought 千 and 干 can be considered similar...


皮 彼 波


pi, hi, ha?


  皮 kawa
  彼 kare
  波 nami


The advantage though is that even if you don't know the word, you can guess the meaning if it contains familiar character(s). Also except for the first one, your examples are easy to distinguish.


Yep and no. That's a common misconception. Like Chinese, old Japanese went through a systematic simplification of its kanji 100 or so years ago. The rules were decided a priori to make the job easier. This meant that several kanjis ended up looking somewhat similar even though they have nothing to do with each other. That's why comparing kanji can work but it can also backfire and lead you astray.


Words in English are built from characters we can pronounce individually and this extends to words so whilst you might not get it perfect you can at least guess the sound of an English word. This isn't the case for characters in Mandarin or Japanese, no sound is encoded in the strokes, the equivalent of a character.


Actually English spelling is pretty complicated and completely different from pronunciation. Why on Earth you Englishmen write "a" (as in skate) but pronounce it as "ei". And why you have so many versions for each vowel and why you have 3 letters for "k" sound and 2 letters (i, y) for the same sound.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: