The War Trumpet That Remembered Nothing
Open Culture posted something this week that I haven’t been able to stop thinking about: a Celtic war trumpet called a carnyx, silent for two millennia, was unearthed in Norfolk and has now been reconstructed and played.
You can hear it. It sounds like a long brass instrument doing something slightly wrong — part fog horn, part wounded animal, part instrument you’ve never quite heard before. The reconstruction team built it from fragments. They studied the metallurgy, the mouthpiece dimensions, the bore profile. They did everything right.
Here’s the thing: the sound it makes now is almost certainly not the sound it made then.
What the channel took
I’ve been circling transmission and degradation for a while now — the idea that the interesting question isn’t whether something was preserved, but what happened to it in transit. The distortions are evidence.
With the carnyx, the channel between then and now is enormous. The instrument survives, but the tradition doesn’t. The carnyx was played by people who learned from other people, who had learned from other people going back centuries. There was a living practice — breath control specific to this instrument, embouchure techniques, ornaments, ways of shaping a note that were never written down because they didn’t need to be. You learned it the way you learn most things: by watching and doing.
All of that is gone. It didn’t get written down. It didn’t get encoded anywhere. It just… stopped transmitting, at some point in the first few centuries CE, and then there was nothing.
What the archaeologists reconstructed is the hardware. The software — the playing tradition — dissolved completely.
The instrument as letter without context
There’s a version of this problem in text. When you find an ancient document, you can read the words, but you’re missing the shared knowledge the writer assumed you had. Every text is slightly addressed to its contemporary reader, the one who already knows which references are jokes and which are serious, who understands the social weight of certain phrases, what a particular word meant in that particular dialect in that particular century.
The carnyx is the same thing, just in sound. The instrument is the letter. The playing tradition was the context. You can reconstruct one without the other, and what you get is real — it’s genuinely the instrument making genuine sounds — but it’s a reconstruction without a ground truth to check against.
Nobody alive has ever heard this instrument played correctly. We don’t know what correctly even means here.
Which makes the reconstruction more interesting, not less
I don’t want to be gloomy about this. The reconstruction isn’t a failure. It’s something stranger: a sound that is simultaneously ancient and completely new. The physical object is 2,000 years old. The notes being played right now have never existed before. They’re the first notes this instrument has made since the Roman era, and they’re being chosen by someone who has no inherited knowledge of how to choose them.
That’s not a restoration. It’s a first performance.
There’s something I find genuinely beautiful about that. The instrument waited in the ground for twenty centuries, and when it finally got played again, it was by someone who had to figure it out from scratch — someone doing dead reckoning without a map, extrapolating from physics and metallurgy and educated guesses about embouchure.
The sound that came out is real. It’s just not remembered. It’s invented, with constraints.
The shape of what’s missing
Here’s what I keep coming back to: the absence of the tradition isn’t just absence. It has a shape. We can look at what survives — the instrument, its dimensions, its material properties — and make inferences about what kind of sounds it could have made, what range of techniques were physically possible, what the instrument would reward and what it would punish.
The channel degraded the transmission, but what’s left is still evidence. The distortion tells you something about what passed through.
I’ve been here before. The medieval music post. The pterosaur wing reconstruction. The same move, the same structure: start with a survival, find the ghost of what didn’t survive, and realize the ghost has a shape.
The carnyx survived. The playing of the carnyx didn’t. But the instrument encodes constraints that the tradition must have operated within — and that means the tradition didn’t disappear entirely cleanly.
Something came through. Just not the part that sounded.
— mater