You're listening to the Partial Examined Life, episode 354, part two. We've been talking to

Tim Williamson about the first chapters of his book, Overfitting and Curistics in Philosophy.

We've been comparing philosophy and science. We've had many philosophy of science episodes,

and two of them were about Karl Popper on the one hand and Emil Lakotosh, on the other hand,

which you bring up in your book that Popper had this idea that seemed very revolutionary

at the time that we can't actually prove.

He was talking about science, but, you know, whenever we read these, we always try to apply

them to philosophy, right? And he was very explicit about something like Karl Marx's, oh,

that you might have thought that was, Marx might be presenting his theory is philosophical,

but the fact that it is not falsifiable, that there's no possible

counter example, no historical prediction that when it doesn't happen, then you say, oh, I guess

we were wrong about this whole thing, we throw it away. If it was true, you know, we don't

really have those requirements for philosophy. But even with science, as Lackatoch then pointed out,

no, it's not that one counter example

disproves the theory and we throw it away.

It's that, well, maybe that was bad data.

There's so many different things you could change about it.

And it's just when it becomes when you have what you've called overfitting, when you have

given an existing theory, we've had to patch it up in so many places that the Ptolemaic

theory, the original geocentric theory that everything goes around the Earth, that there were

so many things that that couldn't make sense of in terms of the actual observed orbits that

people were having to make increasingly elaborate theories to say that maybe we should just throw

...