How were Chinese characters used to record Japanese sounds before kana developed?

Before the development of the kana syllabaries, Japanese writers used a system known as man’yōgana to record the Japanese language phonetically.1

Emerging in the 8th century, this method utilized Chinese characters based on their pronunciation rather than their original ideographic meaning. For example, to write the Japanese word yama ("mountain"), a scribe would select one character that sounded like ya and another that sounded like ma. While this phonetic application appeared in earlier texts like the Kojiki, it was most extensively employed in the Man’yōshū, from which the system derives its name.12

By the 9th century, these characters were simplified into the rounded hiragana and angular katakana scripts, which allowed for a more efficient and complete phonetic transcription of Japanese grammar and vocabulary.12