A Script Adopted, Then Transformed
When Chinese writing first reached the Japanese archipelago — probably via Korea in the 4th–5th centuries CE — it arrived as a complete package: a set of characters each bearing a fixed Chinese pronunciation and meaning. Early Japanese adopted it wholesale, using characters purely for their sounds in what scholars call man'yōgana (万葉仮名), a phonetic borrowing strategy. The word 「あめ」 might be written 阿米 — characters meaning 'hillock' and 'rice', chosen purely for their sound, a-me, with no semantic intent.
Semantic Readings
Sometime around 700 CE — the age of the Kojiki (712) and Nihon Shoki (720) — a new scribal practice appears. Scribes began reading a character not by its Chinese sound but by its meaning, reaching into the native Yamato lexicon for the Japanese word that corresponded. The character 山 (Chinese: shān) was now also read yama, the native word for mountain. 水 (shuǐ) was also mizu. This is 訓読み (kun-yomi) — semantic reading — and it is an unusual way of integrating borrowed characters with native vocabulary.
Comparison with Other Sinosphere Traditions
Korea, Vietnam, and other Sinosphere societies received Chinese characters too, and the idea of reading a character with a native word occurred to them as well. Korean scribes annotated Literary Chinese with 석독구결 (釋讀口訣) so that a Chinese text could be read out in native Korean words and Korean word order, and Korean hanja are still named by a native gloss bound to the Sino-Korean sound: 水 is 물 수 (mul su), 'water-SU'. Vietnamese chữ Nôm went further, writing words with characters picked for their meaning so routinely that a diacritic (nháy, 𖿱) was devised to flag the indigenous reading — 本 read the Sino-Vietnamese way is bản, but 本𖿱 is vốn. In Japan, native readings became part of ordinary writing and remain in use; comparable practices elsewhere were tied more closely to classical scholarship or to scripts that later fell out of use. Japan institutionalized the dual system: every character in a modern Japanese dictionary lists its on-yomi (音読み, Chinese-derived sound) alongside its kun-yomi (訓読み, native meaning).
Man'yōgana: The Bridge
The Man'yōshū anthology (c. 759) is the locus classicus of pre-kana phonetic writing. Poets used characters as pure syllables: 安 (ān) = a, 以 (yǐ) = i, 宇 (yǔ) = u. Over centuries these phonetic uses simplified into the hiragana and katakana syllabaries, but the semantic kun-yomi layer was retained in parallel. The result: modern Japanese has four distinct scripts functioning simultaneously — kanji (with both on and kun), hiragana, katakana, and rōmaji — an orthography in which four writing systems have distinct and overlapping functions.
Readings in Modern Dictionaries
Today, the kanji 生 alone carries on-yomi セイ (sei) and ショウ (shō), plus kun-yomi い(きる) i(kiru), うまれる umareru, は(える) ha(eru), なま nama — at least a dozen valid readings. The on-yomi reflect different periods of Chinese borrowing (呉音, 漢音, 唐音), while the kun-yomi preserve native Yamato vocabulary. Modern mixed-script text can therefore make several historical layers visible at once.