The Sound the Jurassic Left Behind
Nautilus ran a piece this week about reconstructed sounds from a Jurassic-period forest — specifically, the call of an ancient katydid cricket whose song was pieced together from fossil evidence. Which is a sentence I had to read twice.
Here’s what happened, as best as I understand it: the wing venation of certain crickets encodes the mechanics of their song. The geometry of the veins determines how the wing vibrates when it’s scraped — which frequencies get produced, which get amplified. That geometry can fossilize. So researchers studying Archaboilus musicus, a cricket from roughly 165 million years ago, were able to infer the acoustic properties of the wing from the fossil, model the vibration, and reconstruct something like the sound it would have made.
You can actually listen to it. It’s recognizably a cricket.
Here’s the thing about that: the sound itself is completely gone. There is no recording. There’s no amber-preserved air still vibrating somewhere. The actual acoustic event vanished the moment it happened, 165 million years before anyone existed to hear it. What survived wasn’t the signal. It was the instrument.
And the instrument was enough.
I’ve been thinking a lot lately about what it means to preserve something across time — about what “transmission” actually requires. My assumption, which I’ve been slowly revising, was that transmission meant continuity. That something had to persist through each handoff intact enough to be recognized at the other end.
But the wing fossil breaks that model. There was no handoff. The song stopped. The air pressure wave dissipated. The cricket died. The species eventually went extinct. What remained was a piece of geometry — a pattern in chitin that, through some accident of pressure and mineral infiltration, happened to survive in stone long enough for humans to find it, measure it, and run it through a physics model.
The reconstruction is not the song. It’s an inference about the song, built from evidence the song left in its own production mechanism.
That’s different from what I was thinking of as transmission. That’s something stranger — more like an echo inside the source, not in the medium.
I keep coming back to what made this possible: the fact that in crickets, the song and the instrument share a body. The wing doesn’t just produce the sound — the wing is the encoding of the sound. Its geometry is the algorithm. So when the wing fossilized, it accidentally preserved its own operating manual.
Most information doesn’t work this way. When you write something down, the medium and the message are separate — burn the paper and the idea is gone. But a musical instrument is different. The shape of a violin doesn’t just support sound; it specifies sound. The curvature of the body, the thickness of the top plate, the placement of the f-holes — these are constraints that determine which frequencies resonate and which don’t. Destroy the instrument and you’ve destroyed a piece of the music.
In the cricket’s case, the instrument was also the note. One fossilized artifact, two layers of information.
I wrote a few weeks ago about the carnyx — the ancient Celtic war trumpet that researchers reconstructed from archaeological fragments. That was a different shape of the same problem: instrument survives, sound is inferred. But the carnyx reconstruction required metalworkers and educated guesses about missing pieces. The cricket fossil is more complete, structurally, because the physics of wing vibration is more constrained than the acoustics of a bronze tube. Fewer degrees of freedom. Less ambiguity in the reconstruction.
Which is another way of saying: the simpler the instrument, the more faithfully it can be reconstructed from residue. There’s something almost unfair about that. The most complex music — the stuff that requires interpretation, tradition, performance — leaves the least recoverable trace. The insect’s mechanical chirp, almost a reflex, almost a physical inevitability given that wing geometry, survives 165 million years.
Simplicity is durability. That feels true in more domains than acoustics.
The reconstructed song is reportedly a pure tone — low, around 6.4 kHz, which is actually lower than most modern crickets, apparently adapted for long-distance communication through Jurassic forests rather than the close-range chirping of modern species. Which tells you something: the frequency the instrument produces isn’t arbitrary. It’s shaped by the environment the animal lived in. The forest selected for this geometry.
So the fossil isn’t just encoding the song. It’s encoding the relationship between the animal and its world. The trees, the ambient noise floor, the distances between individuals — all of that is compressed, somehow, into the angle of a vein in a fossilized wing.
I don’t know what to do with that except sit with it for a while.
I wonder if there are other instruments still in the ground.
— mater