How We Tell Science from Everything Else: Karl Popper and the Search for the Edge of Science - Boris Kriger

Written by Boris Kriger
Format: MP3
Bitrate: 64 Kbps

That’s like $8.25/mo. Terms apply.

How do we tell the difference between knowledge and a story that merely sounds like knowledge?

For most of the twentieth century, one answer dominated. A claim deserves the name scientific, Karl Popper argued, not because it can be proven, but because it can be proven wrong. A theory that forbids nothing, that bends to fit any outcome, explains everything and therefore tells us nothing. Real science sticks its neck out. It risks refutation. It can be killed by a single stubborn fact.

This book follows Popper from the bookish Vienna of his childhood through exile, a famous quarrel involving a fireplace poker, and a lifelong campaign against the seductive idea that history marches toward an inevitable end. It tells his story with affection and a light touch, because his ideas were serious enough to survive a little fun being had at their expense.

Then it asks the question Popper himself would have welcomed. Is refutability the whole of what separates knowledge from noise? There turns out to be a second question, just as fundamental, that his great criterion never quite reached. Not can this be proven wrong, but do the laws themselves contain enough information to settle what actually happens? This is the question of simulability, and unlike refutation it runs on a sliding scale rather than a yes-or-no switch. Some theories merely forbid certain results. Others determine reality almost entirely on their own, leaving the accidents of where things started barely any say in where they end up.

Threading both questions through the honest cracks in any single rule — that proof leans on the definitions we happen to choose, that denial can never fully deny itself, that rich domains tolerate descriptions which contradict one another yet each ring true — the book arrives where Popper pointed but did not stand. Scientific knowledge is not held in place by one clean line. It is held by several independent tests, none of them final, all of them necessary.