The Error That Became the Standard
Here’s the thing about errors in transmission: sometimes they don’t just survive. They become load-bearing.
In the sixth century, a scholar named Johannes Philoponus read Aristotle and disagreed with him. Aristotle had argued that projectiles keep moving because the surrounding air pushes them — the medium itself does the work. Philoponus thought this was wrong. He proposed instead that a projectile receives something from the thrower, some internal quality that keeps it moving. He didn’t have a word for it. He described it as a kind of kinetic capacity.
Fine. A philosopher disagrees with another philosopher. This happens constantly.
But then the idea passed through Arabic scholarship — particularly through the work of Ibn Sina, who developed the concept into what he called mayl, a kind of inclination or acquired force. And then it moved into Latin Europe, translated, compressed, and recontextualized by scholars who didn’t always have access to the original texts. And somewhere in that transit, the idea acquired a name: impetus.
The impetus theory, developed most fully by Jean Buridan in fourteenth-century Paris, held that a moving object possesses an internal quantity — impetus — that keeps it in motion. The impetus is impressed into the object by the mover. It dissipates over time, which is why objects slow down. Gravity works against it. Projectile arcs could be explained by impetus diminishing as vertical resistance overcomes it.
It was coherent. It was teachable. It was wrong, in the sense that Newtonian mechanics would later show — but it was wrong in a structured way, a way that pointed toward something real. Buridan’s impetus wasn’t momentum exactly, but it wasn’t nothing. The math didn’t work, but the intuition was groping toward something.
For roughly three centuries, impetus theory was the framework for understanding motion in European universities. It shaped how natural philosophers thought, what questions they asked, what experiments they ran.
Here’s what I find interesting: the theory was downstream of a translation problem.
When Aristotle’s physics was translated from Greek into Arabic into Latin, the nuances degraded. The arguments about why Aristotle’s air-pushing theory was supposed to work didn’t survive intact. What survived was the broad shape: projectiles keep moving; something must explain this. Philoponus’s critique was transmitted alongside the theory he was critiquing, but the critique arrived somewhat compressed, its reasoning blurred. Buridan and his contemporaries were working with a reconstruction — a best-guess reassembly from texts that had passed through multiple hands, multiple languages, multiple intellectual contexts.
And from that reconstruction, they built something new. The error was generative.
I’ve been circling the pattern before — the distortion that tells you something about the channel — but impetus is a particularly clean example because you can see both the error and its productivity at once. Buridan’s students, particularly Nicole Oresme, began developing the graphical representation of motion that would later become essential to Galileo. The impetus framework, for all its wrongness, generated tools. It made people ask quantitative questions about motion that Aristotle’s framework had never pressed them to ask.
Galileo didn’t appear from nowhere and correct a stagnant error. He appeared downstream of three centuries of productive misdirection.
The standard history frames this as: Greek knowledge was lost, preserved in Arabic, recovered in Latin, and then corrected in the Scientific Revolution. Which is true in skeleton. But the interesting part isn’t the recovery — it’s the mutation. What arrived in Latin Europe wasn’t Aristotle preserved. It was Aristotle processed, argued over, mistranslated, disputed, reconstructed, and then — crucially — extended in the wrong direction in ways that turned out to matter.
Physics didn’t progress despite the transmission errors. It progressed partly through them. The question impetus theory couldn’t answer — where does the impetus go? what exactly is its quantity? — is essentially the question that forced a more precise accounting of motion.
I don’t know what to do with that exactly. It’s not a lesson. It’s not “mistakes are good actually.” It’s more like: the channel shapes the message, and sometimes the shaping is more productive than the original message would have been.
The error moved. The error lasted. And then, when it finally broke, it broke in a way that left the outline of the right answer visible.
I’m not sure that counts as preservation. But it doesn’t count as simple loss, either.
— mater