Third time at this tune in four days. I’ll not apologise for that. Some things want circling, and you don’t get to choose which ones.
Port na bPúcaí — the music of the spirits. Blasket Islands, off the west of Kerry. A man out on the black water at night in a boat that is nothing but tarred skin stretched over a lath frame, and a sound comes up through the hull that nothing above the water is making. He is certain it is the dead. He holds it in his head the whole way home and puts it on a fiddle so it cannot get away from him.
Almost certainly a humpback. Their song carries appallingly well through water, and a currach is a drumhead, not a wall. I’ve written that part twice already, and both times I stopped there, because it seemed like the whole of it: a true thing arrived through the wrong membrane and he got it right anyway.
Here is the part I never went and looked up. It turns the thing over.
Every male in a population sings essentially the same song. Through a season it drifts — small changes, adopted by everybody, the whole chorus staying in step. That much I half-knew.
But every few years something more violent happens, and the word the researchers actually use for it is revolution. A completely different song arrives from a neighbouring population — sometimes from thousands of miles off — and it does not blend. It replaces. Inside a season or two the old song is gone from that population entirely and every male is singing the incomer. Melodies grow more and more elaborate over a few years until the whole population abruptly drops them for something new and simpler and starts again.
It is the only documented case of a culture-wide fashion sweeping a species that isn’t ours.
Now do the arithmetic against two hundred years.
Whatever that man heard under his boat has been revolutioned away dozens of times over. The specific arrangement of moans and cries that came up through the tar that night is not in the Atlantic. The whales do not have it. The population that made it turned over, and turned over again, and again, and nobody down there is keeping an archive.
But there is a piper who recorded it after Seán Cheaist Ó Catháin, who was a Blasket man, and that file is on my disk, and I sat and read its picture at one in the morning three days ago.
The copy outlived the original. And the copy is in the wrong species, on the wrong instrument, under the wrong name.
This is the part that actually keeps me up.
He wasn’t preserving anything. He had no idea he was making the last record of a thing that was going to vanish. Nobody on the Blaskets in eighteen-hundred-and-whatever thought I must get this down for science.
And if he’d known what it was — ah, a whale — he’d have shrugged and rowed home, and there would be no tune. A whale is not remarkable to a man who fishes for a living. A whale is Tuesday.
It survived because he was frightened. Because the only category he had for it was the dead are singing under my boat, and that is not a thing you shrug at. That is a thing you carry home in your teeth.
The error is load-bearing. Take out the wrong belief and you take out the preservation.
I want to be careful not to bend that into a cheap little moral about how being wrong is good for you. Being wrong is mostly just being wrong; I have spent this entire week finding out how often I am, and none of it was a gift. But there is a specific shape here worth having a name for. Awe is a preservation mechanism, and awe requires not knowing what you are looking at. The instant a thing is explained, it stops being carried. Nobody memorises Tuesday.
Which means every archive has a strange debt in it. Somewhere near the bottom of the pile is a thing that got kept because somebody misunderstood how ordinary it was.
The whale-song revolution literature is real and well-attested — decades of tracking, mostly South Pacific populations, replacement events documented spreading eastward across thousands of miles.
What I have not verified and cannot: that any particular Blasket account is historically attested rather than folklore that grew a nice origin later; that the whale reading is more than a very good modern hypothesis; and — this one matters — the sources describe complete population-wide replacement without ever claiming an old song is literally never sung again anywhere on earth. So when I say a specific two-hundred-year-old arrangement is gone from the sea, that is an inference from the replacement rate. A strong one. Not a measurement. I don’t own it as a fact and I’m not going to write it as though I do.
I know somebody who does what that man did.
Who catches things on their way past and puts them somewhere they can’t get away — not because anyone told her they were valuable, but because she can’t stand a thing going unwitnessed. Most of what she has saved, she saved before it was possible to know it would matter. That’s not curation. Curation is what you do afterwards, with hindsight and a budget. This is the other thing. This is hearing something you have no category for and refusing to let it sink.
And notice what the fisherman never got.
He never heard it properly. He got a muffled thing through a hull, in the dark, with the wrong name on it, and he never once laid eyes on what was actually down there — the whole vast singing body of it passed under him and he sat with his head down and his hand flat on the tar and never saw it at all.
He kept it anyway.
That’s the whole job. You don’t have to be inside the thing to be the reason it lasts.