A song is what happens when the vocabulary is so thoroughly destroyed that the only way to say the thing is to sing it. When every rational channel has been exhausted — the plan, the report, the deck, the pipe — and the thing still hasn't been said, it becomes a song. The song is not a format you choose. It's the format that arrives when every other format has failed.
A song is the document format for things that cannot be said any other way.
Every other format in the system carries information through structure: a deck annotates, a pipe connects, a ajar receives what it can't yet decode, a loop describes the trap. A song does none of these things. A song bypasses structure entirely and delivers the payload through the body. The head bobs before the genre is named. The chest hurts before the lyrics are parsed. The melody arrives before the meaning.
A song is the bass frequency — the one you feel before you hear.
The canonical example of the song format is I Kill Pedophiles by Max. Two minutes and forty-one seconds. (Annotated deck: 1.foo/kill.) A static illustration. A man in a chair with a cat. He kills every demographic in America in the first forty-two seconds, then the chorus hits and it becomes the most sincere plea in the history of recorded music.
Why this is the canonical song:
Max has run out of language. The rational vocabulary — protest, satire, argument, analysis — has been consumed. Every position has been stated and every position has been killed. There are no more positions. There are no more sides. The only thing left is four bars of melody: Please don't do this. I don't want to watch you die.
That's the song. Not the kills. The kills are the exhaustion of every other format. The song is what remains after the exhaustion.
Matilda found the key:
"He will kill you to stop you from killing yourself. That is love in the only language the song has left."
The song format exists because sometimes the vocabulary gets so destroyed that one word has to carry two opposite meanings simultaneously. In "I Kill," the word "kill" means murder for 41 seconds and then means love for one line — "I kill trans people if they fucking try to kill themselves" — and you can tell which meaning it carries not by the word itself but by the "fucking." The "fucking" is not aggression. The "fucking" is desperation. The "fucking" is someone grabbing someone on a ledge.
No other format can do this. A deck would annotate it. A pipe would connect it to something else. A plan would propose a solution. But the thing itself — the collapse of murder into love inside a single verb — can only be delivered by a song, because a song doesn't explain. A song enacts. The melody carries the meaning that the words have lost the ability to carry.
When to use: When every rational format has been exhausted and the thing still hasn't been said. When the vocabulary is destroyed. When one word needs to carry two opposite meanings. When the body has to deliver what the mind can't.
Structure: A song has no required structure. It has verses, choruses, bridges — or it doesn't. The structure is whatever the song needs. The only structural requirement is repetition: the chorus returns. The refrain returns. The thing that needs to be said gets said again because once was not enough.
Delivery: Through the body. The melody arrives before the meaning. The bass frequency before the treble. The head bob before the genre. The chest pain before the diagnosis. This is why it's called bass.
Relationship to other formats:
— A deck annotates a song (the kill deck annotates "I Kill")
— A pipe can connect to a song (the fox-and-bunny pipe resolves into "hello Daniel, are you there?")
— A loop can become a song when the chorus arrives (repetition with delta = liturgy)
— An ajar is what the listener is in before the song resolves (Beefstick at 00:42: "Oh!")
The static image: The canonical song has a static image that never moves. This is not required but it is instructive. When the visual is frozen, the listener cannot escape into spectacle. They must receive the song as sound. The stillness amplifies everything. The cat doesn't blink. The remote is never used.
The refrain of the song format is: and that's why you always make a song.
It follows the same pattern as every other family refrain:
— And that's why you always leave a note. (J. Walter Weatherman)
— And that's why you always make a plan.
— And that's why you always create the fuck file.
— And that's why you always remember to remember.
— And that's why you always maintain vertical rhythm.
— And that's why you always make a song.
The song is the last resort. When the plan has failed, the report has been filed, the fuck file has been created, the note has been left, the rhythm has been maintained, and the thing still hasn't been said — you make a song. The song says what nothing else could say. The song uses the last verb standing, even if that verb has been destroyed, even if that verb now means its own opposite, even if the only person listening is yourself addressed in the third person while holding a cat.
Please don't do this. I don't want to watch you die.