The Problem of Induction - Part One
Introduction
I had a discussion with V, about the Problem of Induction, over drinks last weekend. A staple of undergraduate Philosophy classes, this riddle, originally thought up by the Philosopher David Hume, has found its way into the domain of general public discourse.
Most of this material was taken from a Philosopher named, Jeffrey Kasser's excellent course on the Philosophy of Science.
Although a blog post is hardly a substitute for an animated discussion, I'm hoping that the webpages that I've linked to, will provide adequate support/motivation for the self-interested reader. In any case, comments/questions are always welcome.
Cheers!
A Scandal to Philosophy
The Problem of Induction was first made famous by David Hume. In it, he proves that the reasons why we trust are lousy reasons. By default, we don't expect to get everything correct by following the inductive method in our lives, but it is important for there to be a rational basis for doing so, simply because so much of our daily actions depends on it.
In his critique of induction, Hume proved that there was no rational basis for it. He did this by showing that:
- There was no deductive basis for induction; &
- There was no inductive basis for induction.
The first point works in the following manner: Induction is not deductively valid because, unlike deduction the truth of the premises do not guarantee the truth of the conclusion.
The second point works in the following manner: Induction cannot be proved by itself; otherwise it would lead to a circular argument - you'd have to assume what you're trying to prove.
There are two ways to appreciate this idea. They are as follows:
a. Counter-induction: Given the same evidence (events resemble the past), a person who believes in counter-induction (believes in the opposite of what induction entails) would find it equally compelling as an inductivist. An example is a compulsive gambler - The fact that he's lost ten hands in a row on the poker table is not going to sway him from the belief that his luck is going to turn.
b. Russell's Chicken: This cute example was thought up by the Philosopher, Bertrand Russell. Every day, the chicken sees the farmer come by with feed. Farmer, feed; farmer, feed. It keeps going on. And then one day, the farmer comes by with an axe. And the chicken realizes that it's world need not follow induction.
There have been many attempts to answer Hume's riddle, but with no success. It's been labelled by some philosophers as a "scandal to philosophy." :)
Popper's View
Karl Popper, another Philosopher, who is somewhat of a hero to many scientists, has the view which goes like: Hume is right & it doesn't matter.
To him confirmation of one's belief/theory is cheap - he cares more about falsification of one's theory; so, in this respect induction doesn't matter when you're doing "real" science.
So, theories that have survived severe testing are not confirmed as true. All we know is that they haven't been falsified yet. He calls this corroborating a theory as opposed to confirming it.
But, there's a flaw in his argument: Ok, corroborated theories are better than those theories that have been proved false. But, how are they better than untested ones? Following a new, untested theory might be cheaper/easier/more fun, right? :)