> Surely it doesn't really start until the first note is played?

That question itself is built on a radical assumption. Example:

Just skimming, it looks like 38 out of 48 of the fugues from Bach's WTC Books 1 & 2 begin with rests followed by several beats worth of melodies in the first measure[1]. If you think the piece only really starts when the first note of the melody is played, then you've got problems. Either:

1. You hear the first articulation of the melody as the downbeat. But that would mean the first measure is in a disparate (and probably irregular) meter. Radical!

2. You hear the incomplete first measure as an anacrusis, or a kind of unaccented lead-in, to the second measure: OMG even more radical!

You can use accent patterns on a modern piano to play any of these fugues using either of these methods, and it will sound silly to non-silly keyboard players.

What's more, non-silly keyboard players do feel the pulse for the first downbeat of these pieces when they perform. Most of them will even inhale before the downbeat, as if they are somehow singing the melody through their fingers.

Finally, lots of music begins with rests: not just conservative cases like Bach, but progressive cases like the beginning of Beethoven's Fifth, and of course the radical cases like Cage's.

This leads to one of the things I love about Cage's music: it compels criticisms that reveal fundamental misunderstandings about music. E.g., you don't buy the radical case-- perhaps reasonable. But you then speculate there are no cases-- which is at odds with common musical practice.

If Cage's music did nothing but compel these questions it would be worth its weight in pine nuts.

Edit:

1: Bach does this because nearly all the fugues have three or more independent melodies singing at the same time. If they are all singing on every downbeat it can quickly sound really clunky and predictable.

> 2. You hear the incomplete first measure as an anacrusis, or a kind of unaccented lead-in, to the second measure: OMG even more radical!

Why is this radical at all? This is exactly how most humans perceive it: as a lead-in to the second (I might even argue first) measure. It's very strange to me to say that most humans are supposed to understand the piece to have started before any sound is played. In fact that's quite preposterous: play a song that starts with a rest to 1000 people and ask them to gesture as soon as the song starts, and every single one of them will gesture on the first note played. How are they supposed to perceive the song to have started any earlier than that? A song "starting with rests" is written that way to make it understandable to the performer who is reading the notation. It's a purely notational thing. The notation is not the song, the sound is the song.

> If you think the piece only really starts when the first note of the melody is played, then you've got problems. Either:

> 1. You hear the first articulation of the melody as the downbeat. But that would mean the first measure is in a disparate (and probably irregular) meter. Radical!

> 2. You hear the incomplete first measure as an anacrusis, or a kind of unaccented lead-in, to the second measure: OMG even more radical!

Why? Why can't you just say the piece starts partway through a bar, and we notate that with a rest for convenience? Just as when a piece ends partway through a bar we would generally accept that it ends when the last note ends (and while we might notate that as being a full bar in the case of a long held note, we don't always play it that way), not after some trailing rests, and we wouldn't consider this as being some kind of radical accented thing.

> This leads to one of the things I love about Cage's music: it compels criticisms that reveal fundamental misunderstandings about music. E.g., you don't buy the radical case-- perhaps reasonable. But you then speculate there are no cases-- which is at odds with common musical practice.

Love this thought. You disagree with an extreme interpretation, do you take the exact opposite? If not, up to where do you go?

This idea is applicable to so much