On “proof” in science

By Peter Bokulich

Phil Plait (the “Bad Astronomer”) deals with a denialist who thinks only a bad scientist would claim there’s no place in science for “proof.”

A few quick points that tie into lecture:

1. When logicians and mathematicians speak of a “proof” what they have in mind is a demonstrably valid argument form. Keep in mind that validity is distinct from soundness.

    

2. Typically, when a laypeople speak of “proof” then have in mind a set of evidence that should produce a degree of confidence in the truth of some claim. The question of just how much confidence is warranted by the evidence is the crucial question here, and is one that is left unaddressed by the layperson’s usage.

    

3. If “proof” is supposed to imply certainty in the truth of some conclusion, we will never have a “proof” of anything.

    

4. The legal notion of “proof beyond reasonable doubt” is arguably out of place in science.

    

5. Our thinking and speaking about these matters will be clearer if we employ the concept of degree of uncertainty and we assess the risks in light of the uncertainty.