Tuesday, April 24, 2007

Knowing what's so

How do we know something is true?

In many cases, we don't. We may have to make good decisions based on partial data, or we'll be left behind.

Sometimes I hear someone wish they could have the certainty one gets from mathematical proofs in their own field.

Papers such as Social Processes and Proofs of Theorems and Programs by De Millo, Lipton, and Perlis have served me as healthy reminders that mathematicians don't have it easy, either. Even for them, the process of knowing something is heavily a social process, even as outsiders may view theirs as an ironclad, logical process.

That's not to say we should be sloppy in our explorations, investigations, and research; we shouldn't. It just says we should have a realistic understanding of the way things are really done.

As an aside, there's an interesting and humorous description of one particularly problematic use of randomized control trials that some of you might enjoy.



Post a Comment

<< Home