Warning: Shakuhachi inside baseball ahead.

So one of the genres of music typically played on the shakuhachi is honkyoku 本曲: mostly solo pieces, usually without strong rhythm, handed down from komusō culture. Kyorei 虚鈴 ("Empty bell", etc.) is one of the "three classics," traditionally considered the oldest surviving honkyoku, and has been transmitted in many different versions.

Romei at shakuhachi blog Nichiyō komusō made an interesting post not long ago about three versions of Kyorei. His diagram shows a summary, in shakuhachi notation (black text = low register, red text = high register, blue text = low or high register), of the piece in three versions: Seien-ryū 西園流, Taizan-ha 對山派, and Jin Nyodō 神如道's version (i.e. the Fudaiji version).

Now, the Taizan-ha version of the piece is based on the Seien-ryū version, edited into its present shape by Taizan-ha founder Higuchi Taizan 樋口對山. This editing seems to consist mostly of adding repetitions (e.g. as in the first row) and moving the first few phrases to the high register. This is relatively uncommon in the honkyoku repertoire, in which pieces tend to start low, go high, and then come back down to low again. (The "mountain" structure.) So why did Taizan do that?

Romei argues that both the repetitions and the high-register start were a way of reinforcing the "Kyorei sound world," the chief characteristic of which is eternity. The repetitions undermine the passage of time, eroding what little motivic direction the Seien-ryū version originally had, and the "start high" approach is a way of implying that the song was already in progress when the player picked up the flute. What is played, and heard, is just the section of Kyorei that happens to be manifest in our world; and the ending, too, implies a continuation in another place.

I'm not in a position to judge the validity of this (I've never even spoken in person with a Taizan-ha player), but I enjoyed the argument.