In the early days of paleontology, scientists were doing something gloriously human: staring at a pile of ancient bones and confidently inventing monsters.
Take Richard Owen and the first dinosaur reconstructions at the Crystal Palace Park in the 1850s. The dinosaurs look like lizards that swallowed anvils. They’re squat, elephantine, vaguely embarrassed. Why? Because Victorian scientists only had fragments—teeth, vertebrae, a femur here and there. They filled in the blanks using what they already knew: iguanas, crocodiles, rhinos. The unknown was interpreted through the lens of the familiar.
(one of the earliest and worst reconstructions)
Then there’s Hydrarchos, a “sea serpent”,
Stitched together from multiple whale fossils by showman Albert Koch. It fooled crowds because people wanted it to be real. The bones were genuine. The narrative was not.
Even honest mistakes were spectacular. Early reconstructions of Iguanodon famously placed its thumb spike on its nose like a rhinoceros horn. For decades, museum visitors absorbed a completely wrong animal as scientific truth. Not because scientists were foolish—but because inference outran evidence.
Science corrected itself, slowly. More fossils appeared. Comparative anatomy improved. The scientific method—the disciplined practice of testing hypotheses against new evidence—started sanding down the imaginative excess. Each bad reconstruction was a scaffold. Wrong, but useful. The errors taught scientists how not to see.
History works the same way.
We have fragments: a few surviving letters, biased chronicles, ruins, coins, laws carved into stone. The overwhelming majority of human experience—daily life, private thoughts, entire cultures—has dissolved. So we reconstruct.
We build narratives out of shards.
Like early paleontologists, historians use analogy. If one medieval village kept certain records, maybe others did something similar. If a king wrote propaganda, perhaps he ruled a certain way. Gaps get filled with educated guesses. The unknown is interpreted through the familiar.
Sometimes we place the thumb spike on the nose.
We call it “the fall of Rome” as if a single dramatic collapse explains centuries of transformation. We describe entire eras with tidy labels—“Dark Ages,” “Renaissance”—as though history naturally organizes itself into chapters. These are intellectual skeletons we’ve assembled from scattered bones.
That doesn’t mean history is fake any more than dinosaurs are fake. It means reconstruction is unavoidable. The key distinction is between hypothesis and certainty.
Early fossil science teaches a humbling lesson:
Confidence should scale with evidence. When we have three bones, we should speak lightly. When we have a full skeleton, we can stand taller—but even then, soft tissue, color, behavior, ecosystem dynamics are still partly speculative. A fossil gives structure. It does not give life.
But reality is lumpy. History is more like a box of mismatched bones from different animals, and we are trying—earnestly—to assemble a creature that may never have existed exactly as we imagine it.
There’s something beautifully sobering about that. Knowledge is not a statue. It’s a sketch constantly being redrawn.
The deeper lesson isn’t cynicism. It’s intellectual humility. Every reconstruction—of a dinosaur or a civilization—is a working model. Some models are robust. Some are embarrassingly wrong. Most sit somewhere in between.
We are always assembling bones in the dark. The question is whether we remember we’re doing it.