SONG: 14.8 Million


“14.8 Million” [Download]

ARTIST: grant.

SOURCE: Nature 14 Dec 2022, “Missing data mean we’ll probably never know how many people died of COVID,” as used in the post “More people died of COVID-19 than we know. Or ever will know. Because data is missing.”.

This is, shall we say, one the more “experimental mode” songs. I spent a month, more or less, playing around with a hang drum VST, making a kind of ambient soundscape thing with some overlapping loops (I hesitate to call them polyrhythms), and no idea really what to write words about.

I’m really sort of tired of covid news – on the one hand, it’s super important (and we’re still getting it wrong), but on the other hand, it’s a topic that has sort of overtaken everything else in the world of science publishing, and as far as narratives go, they’re all basically the same. I want more corners to be turned, more breakthroughs, more triumphs over adversity at long last.

Instead, the newest sort of story is this: we will never know exactly how many people we have lost. We still don’t know how to do the simplest thing, which is to count.

All I knew when I settled on this topic was that I didn’t want it to rhyme, and I wanted it to be sort of a mess of different voices, overlapping. I didn’t sing on this at all. There is no melody. This is seven or so text-to-speech websites using different voices to repeat the same words. It starts with a child and ends with a Chinese woman.

The math is a back-of-envelope calculation. I like doing that with big numbers, because 20 people is a tragedy, 200 is unfathomable, but 2 thousand and 2 million are both equally statistics, you know? I have to make that kind of figure concrete. So here’s where I start: If the figure of 14.8 million uncounted people is correct, then to read out all of their names, at one second per name, would take five months and three weeks.

But we can’t do that, can we. We can’t even get all of their names on one list. Someone, somewhere knows each name, but collectively, we can’t do that. That means these are each private losses. The loss is of historic proportions, but it’s not part of history.

There is a sense where I was putting this together and thinking of the songs of the dinosaurs that we’re only now reconstructing, because they seem not to have had larynxes, and that kind of soft tissue isn’t fossilized anyway. But I didn’t want to insert anything as potentially corny or monstrous as dinosaurs into this, and I didn’t want to flirt with the idea of mass extinction either, since that’s a totally different order of magnitude.

These songs that are lost, they are our own.


In five months and three weeks,
We could read every name
One name per second
But we would need to stop to breathe
And we would need to remember

14.8 million songs were silenced
And we have not remembered
14.8 million songs, we can’t remember.
We won’t remember.