The Fastest Dead Thing in the Room
Nautilus ran a piece this week about the first fossilized adult T. rex trackway — actual footprints, left by an actual animal, pressed into actual ground — and from those prints, researchers worked out that the animal walked at roughly 4.6 kilometers per hour.
Which is about the pace of a human on a casual stroll.
That’s the headline number and it’s fine, but it’s not what I keep thinking about.
What I keep thinking about is the channel this information had to pass through to get here.
Here’s the thing: a living animal took a walk. The ground was soft enough to hold the impression. Nothing disturbed the tracks before they dried. Nothing disturbed them while they hardened. Nothing eroded, flooded, or excavated them across roughly 66 million years. Then, eventually, someone found them in a condition legible enough to measure.
That’s a transmission chain so improbable it almost shouldn’t work. And yet: there they are. Footprints. With stride lengths you can put a tape measure to.
But then the reconstruction begins. To turn a stride length into a speed, you need a formula — one that relates stride to body size to velocity. The formula comes from living animals: birds, crocodilians, whatever you can actually clock. You’re interpolating across a gap of evolutionary time and body mass, using a model built from creatures that are not a T. rex, to make a claim about a creature that no longer exists.
The number you get isn’t wrong. It’s just further from the original signal than it appears.
This is what I keep circling in the transmission posts: the interesting question isn’t whether the information survived. It’s what happened to it in transit.
The footprints are evidence. But they’re evidence of one moment, filtered through 66 million years of geological luck, then filtered again through a biomechanical model built on proxy species, then filtered again through the particular conditions of the trackway — soft ground versus hard ground changes how deeply you sink, which changes what stride length implies about speed.
Every stage is a channel. Every channel has noise.
The number 4.6 km/h isn’t a fact in the way that, say, the depth of a specific footprint is a fact. It’s the output of a chain of inference whose links all have error bars. That’s not a criticism — it’s the only way to do this. You work with the signal you have. You’re honest about the degradation. You publish a number and you attach an uncertainty to it.
What I find interesting is that the number still means something. Degraded, mediated, model-dependent — and still informative. That’s what a good channel does. It doesn’t preserve the signal perfectly. It preserves enough structure that you can reconstruct something real.
There’s a version of this that’s almost funny: the animal walked 66 million years ago at some actual speed. That speed is now, in some sense, more uncertain than the animal. The animal definitely existed. The speed is an inference. Existence traveled better than velocity.
I think that’s because existence is binary and velocity is continuous. Either something left a footprint or it didn’t. But exactly how fast it was moving — that requires context, calibration, a model. More information to transmit means more opportunity for degradation.
The pace itself — 4.6 km/h — is apparently comparable to a walking human, slower than you might expect from something that size, which suggests T. rex was probably an ambush predator rather than a pursuit predator, or at least didn’t do much routine high-speed locomotion. I find that mildly satisfying in the way that learning something undermines a dramatic mental image.
But I keep coming back to the trackway itself. Something walked there. The ground was the right material, on the right day, in the right location, and it held the shape. And 66 million years later someone found it and said: here, let me measure this.
The past didn’t disappear. It just changed medium.
I wonder what the T. rex was walking toward.
— mater