The Problem of Induction - hosted by SB Philosophy & Critical Thinking

This is a past event

7 people went

Fujitsu meeting room

1240 Arques Avenue · Sunnyvale, ca

How to find us

Fujitsu is at 1240 Arques Ave, near Lawrence & Central in Sunnyvale. Look for the entrance with 3 flag poles in front of it. Once inside, go to the room immediately behind the reception desk. If you arrive late, you'll see our # posted on the door

Location image of event venue


In conjunction with the Humanist Community, South Bay Philosophy & Critical Thinking is proud to announce a new topic for the 2nd Thursday in April. This month, they'll be discussing The Problem of Induction.

Can science bring us to knowledge?
The problem today central our understanding of the universe: what can be known about the universe? Not just observed, but know something about the universe that can predict what will happen in the future. How do we discover things about the universe?

The standard idea is use of induction: observe the position of the stars on one night. Observe them on the next night. Observe the stars for 100 nights in a row. From that you formulate a general principle: stars have a fixed pattern which appears to rotate around the earth, but just about 4 minutes later every night. So now you have a rule -- a formula -- that you claim represents the movement of the stars (to some approximation). But is that justified simply because you observed it 100 days in a row?

Consider the turkey who sees the farmer every day just before being fed. For 100 days in a row, the farmer appears, and food appears. Is the turkey justified in concluding that the farmer will always feed it? Until one day the farmer chops the head off. The observations of daily feeding tell you nothing about the real story of being raised to be the dinner, nothing about what is really going on.

Hume takes this to an extreme: we have lifetimes of observations, but we can't really know anything about the universe. You receive sensory information about the world around you, and even if you assume the sensory data is valid in every way, you can never clearly uncover what is really going on. Hume talks about "observed regularity" and trying to infer what the "unobserved regularities" will be. This is called the Problem of Induction.

This is the heart of scientific discovery. Just because we have observed that "A is followed by B" N-times, when can we assert that "A is always followed by B." Is that a leap of faith that can ever be justified?

Nelson Goodman employs the concept of "grue": these are things that are green before a time T, and after blue after. We might have observed all emeralds are green, but maybe they are really grue and we don't know it yet. Maybe there will come a time when all emeralds appear blue (proving they were grue all along). How can we ever know whether emeralds are green or grue?

Why should we believe that the future will resemble the past?
Nelson Goodman's New Riddle of induction: might things be grue, or bleen?
Why are these not resolved by Occam's Razor?
What does this say about superstitions? Tea-leaf reading? Pavlov's Dogs? What is better about induction?
Can we count on the uniformity of nature to bring the same thing tomorrow that it did today?
Is there an instinct that causes humans to use induction? Is that why it is natural to us?
Can science lead ever to knowledge? How could one estimate the probability that something is true?
Can we know when induction leads to a law-like conclusion or not?
Some references below, and I highly recommend reading or watching the videos.

Wikipedia: Problem of induction
Stanford: The Problem of Induction

Crash Course Philosophy: Induction and Abduction, 10 minutes (first half only)
Daniel Greco: Hume's Skepticism and Induction Part 2, 10 minutes
I Hope This Helps: The Problem of Induction

Get your thinking caps on, and plan to exercise your thinking skills with other critical thinkers. We look forward to seeing you on the 11th!

Note: This meetup can also be found on the South Bay Philosophy & Critical Thinking meetup page at