Why the Niji script?
March 25, 2018
This document provides background on what led to the development of the Niji script. For more information about the script, see The Niji Script.
Japanese: The most difficult writing system in the world
The Japanese writing system is likely the most difficult one in use today. It uses three separate scripts:
- Kanji: Chinese ideographs, which are used for the roots of most Japanese nouns, verbs, adjectives, and adverbs. The core set that every Japanese child has to learn and that newspapers assume as known contains 2136 characters, but reading any specialized material can require thousands more.
- Hiragana: A syllabic script of 46 core and 7 auxiliary characters, which are used for the grammatical endings and particles in Japanese sentences, for the roots of some words, and to explain the pronunciation of less-well-known Kanji.
- Katakana: A syllabic script of 44 core and 6 auxiliary characters, which are used for words imported from languages other than Chinese, for emphasis, for onomatopoeia, and for various other purposes.
The big problem are the Kanji. First, there are tens of thousands of them, and reading any advanced material requires knowing thousands of them. Second, unlike in Chinese, where one character always corresponds to one syllable, and around 90% of them have exactly one pronunciation and the rest no more than five pronunciations, most Kanji have multiple pronunciations, with no clear correspondence between characters and syllables. For example, depending on context, 美 can be pronounced bi, mi, utsuku, uttsu, o, uma, chu, chū, or can combine with 味 to uma. This makes learning Japanese writing extraordinarily time-consuming. Japanese children spend big parts of their first nine years in school on just the core set of Kanji, and are essentially never done.
Because of its complexity, and the availability of good software, Japanese is well on its way from a writing system to a read-only system: Many Japanese can’t write the less commonly used Kanji anymore, and instead rely on the input methods on their phones or computers to translate from Hiragana or Latin input to Kanji.
The underlying reason for its complexity is that Japanese as a language is very different from Chinese. Japanese does use a large number of words with Chinese roots, in the same way that European languages use many words of Latin and Greek roots. In addition, however, it has a large set of native words, whose pronunciation is unrelated to Chinese, but uses Kanji related to their meaning to write them. A Kanji therefore usually has one or two pronunciations of Chinese origin, plus a larger number of pronunciations of Japanese origin. When it comes to grammar, Chinese largely strings together words without declining or conjugating them, with their function largely determined by word order. Japanese, on the other hand, uses an elaborate system of word endings and particles to express the function of words within a sentence as well as tense, mood, and relative social status of speaker, listener, and the people spoken about. Hiragana has largely taken on the role of expressing the grammar in written Japanese.
The solution for Korean: Hangeul
Koreans used to have the same problem: They similarly used Chinese characters for a language that is very different from Chinese (in many ways, the Korean language is much closer to Japanese than to Chinese). In the 15th century there was an additional problem: Few Koreans had enough time to learn Chinese characters, so most Koreans were illiterate. To solve this, King Sejong in 1446 introduced a new script tailored to the Korean language, Hangeul. It consisted of (then) 28 characters and was so simple that, according to the book explaining its principles, “wise men can understand them within one morning, and even stupid people can learn them in ten days”.
Hangeul is a combination of alphabet and syllabary. At its core is a set of characters called jamo representing the consonants and vowels of spoken Korean. Instead of writing these in a linear sequence, however, the characters of each syllable are combined into a syllable block. For example, the characters for the word hangeul, ㅎ h, ㅏ a, ㄴ n, ㄱ g, ㅡ eu, ㄹ l, are combined into the two blocks 한글. To someone unfamiliar with Hangeul this may look similar to Chinese, but after a few days of practice the jamo and with them the pronunciation are easy to recognize.
Hangeul had a hard time overcoming the resistance of the educated elite of the Joseon dynasty, but today almost all Korean text is written in Hangeul, and Hangeul is widely recognized as a well-designed script.
A solution for Japanese: Niji
The success of Hangeul suggests there could be a similar script for Japanese, a phonemic script that is designed specifically for Japanese and can be learned within a few days. The Niji script is a proposal for just that: It provides characters for the consonants and vowels in modern Japanese, and follows the model of Hangeul in combining the characters of each syllable into a syllable block. Some of the characters are taken directly from Hangeul, some are modified, and some are new.
Why not Hangeul for Japanese?
Hangeul has long been used to explain the pronunciation of Japanese to Koreans. The first Japanese textbooks using Hangeul appeared during the Joseon dynasty; and today Hangeul is used to show Japanese pronunciation in some Korean-Japanese dictionaries as well as to show station names in the Tokyo subway.
However, just like Korean is not like Chinese, Japanese is not like Korean. Hangeul expresses some distinctions between sounds that matter in Korean but not in Japanese, such as aspirated vs. unaspirated consonants. More importantly, it doesn’t express some distinctions that matter in Japanese, such as voiced vs. voiceless consonants, or long vs. short vowels. Using Hangeul for Japanese risks losing information, or confusing speakers by using characters to express different distinctions than they were designed for.
Why not Latin?
Replacing the Japanese writing system with Latin has been proposed numerous times, especially after World War II. There are some arguments against such a move that apply to any phonetic script; those are discussed in the last section. A problem specific to Latin is that the most commonly used transliteration method, Hepburn, does not record full information about the pronunciation, while those that do, such as Revised Hepburn, may use characters that aren’t always supported on computers (long vowels ā, ī, ū, ē, ō).
Why not Hiragana?
Replacing the Kanji within the Japanese writing system with Hiragana would be a more logical choice than Latin, as Hiragana is already an integral part of the system and every Japanese is familiar with it.
However, Hiragana has some deficiencies in representing modern Japanese. Its core character set can represent only 9 consonants directly, while modern spoken Japanese has at least 17 phonemic consonants, 20 if all foreign consonants that can be represented in Katakana are included. For consonants that don’t have a direct representation, various hacks are used:
- dakuten ﾞ is used to turn a voiceless consonant into a voiced one, for example, か ka into が ga.
- handakuten ﾟ is used to turn h into p, for example, は ha into ぱ pa.
- Small letters ゃゅょぁぃぅぇぉ are used to turn one syllable that’s pronounced differently from related syllables into a new series, for example し shi into しゃ sha, しゅ shu, しょ sho.
In addition, the representation of long vowels is inconsistent: A long ō is usually indicated by addingう u, but occasionally with お o.
Finally, a few particles are written with characters that don’t reflect their modern pronunciation: は ha represents the topic marker wa, を wo the object marker o, and へ he the directional particle e.
So, why not Niji?
A number of objections exist against switching to a script like Niji – most of them would apply to any phonemic script, including Latin and Hiragana.
Most importantly, Kanji are an important part of Japanese culture, and there’s a huge body of literature written using them. If people don’t learn Kanji, anything written with them becomes less accessible. The impact can be reduced by transliterating commonly read materials into the new script, but it’s unlikely everything will be transliterated. The situation might become similar to that of the Latin language in Europe: For everyday life, knowledge of it is unnecessary, but if one wants to dig into historical documents or old literature, it eventually becomes a prerequisite.
Another objection is that Japanese has a large number of homophones, words with different meanings but the same pronunciation. Kanji can help to distinguish between them, and are occasionally used even in conversation to disambiguate. Most of the time though, context is or can be made sufficient to disambiguate.
In reading, the alternation between Kanji and Hiragana helps readers find word boundaries and focus on the root of words. The Latin and Hangeul scripts use spaces to identify word boundaries and somewhat help find word roots; for the Niji script the same is proposed.
Well-practiced readers can absorb the meaning of Kanji without sounding out words. However, well-practiced readers of Latin or Hangeul script also recognize entire words without sounding them out – the main difference is that not-so-well-practiced readers can sound out words without having to look up each character in a dictionary.
Objections like these need to be discussed and, in some cases, resolved before Kanji can fade away. However, it seems worthwhile discussing them so that eventually the children of Japan, as well as immigrants, will be able to read everyday materials without first having to learn thousands of characters and their various pronunciations.
For more information about the Niji script, see The Niji Script.