Sam Gentle.com

The testing mindset

I really enjoy reading books by or about scientists, not least of which the inimitable Richard Feynman. I think what is so appealing isn't necessarily the work they do, or any particular discovery or mannerism, but rather a kind of mindset that you don't see much outside of very good scientists: the testing mindset.

What I mean is that there are lots of times when you'll come across something unexpected. Many people won't even notice, because they're not interested or paying attention to something else. Some people will notice, and become curious about the unexpected thing and how it works. An even smaller number will set out to try to learn about or understand the thing. But the rarest response of all is to figure out how to trap this unexpected thing in a web of experiments so it has no choice but to reveal itself. Those people are the ones who get to be great scientists.

A great example is a chapter in Feynman's book called The Amateur Scientist where he mostly talks about ants. He wwas curious about how ants find food and know where to go. So he ran a series of simple experiments involving moving ants around on little paper ferries, setting up grids of glass slides and rearranging them, and graphing their trails on the ground with coloured pencils. He didn't sit around wondering about ants or ask an ant expert, he made specific tests he could run to figure out how they worked for himself. I suspect if ant behaviour had not already been extensively studied, and if he wasn't otherwise occupied with physics, Feynman would have made some significant contributions to the ant field.

I often run into things I don't understand, from the behaviour of some obscure piece of software to my singing tea strainer. I notice, though, that although I'm pretty good at noticing unknown things in the first place, my first instinct is usually to try to learn about them by looking for information somewhere else. That's usually works fine, but what about things nobody knows yet? Looking for answers only works when someone else has already done the work to find them.

It is, of course, way more efficient to learn from the experiments of others than to repeat everything yourself. But if you spend all your time relying on secondhand knowledge you might not build the skills necessary to make new knowledge. The testing mindset doesn't seem like something you turn on and off, but rather a way of looking at the world where you constantly want to poke and prod at the bits that feel funny. So perhaps it's best to do things the hard way sometimes and re-discover from scratch what you could easily learn from a book.

It seems like more fun, at the very least.