Rubberducking and Creativity

You don't need a room full of smart people to brainstorm with--just use a rubber duck.

One of the first things I learned as a new programmer was the technique known as "teddy bear programming". A story about this is in the book The Practice of Programming, and goes something like this: a university help desk center kept a teddy bear, and before students were allowed to bring their problem to a human, they had to first explain it out loud to a teddy bear. The idea is that by the time they finished telling the bear, over half of them had solved their own problem.

But the technique is also known as rubberducking, because it doesn't matter much what you're talking to.

It's the talking that matters. Explaining your problem out loud is often enough to shake things loose in your brain, expose bad assumptions, and cause you to see things in a new way.

Gian's recent comment to the Passion is Infectious blog got me thinking about this. He pointed out how sometimes just having someone come over and take a look at a problem you can't solve is enough to make things happen (and he has a fun speculation on why sometimes two heads are better than one).

The thing is, sometimes it doesn't have to be an Actual Carbon-based Human. Another name for rubberducking is to use the cardboard programmer. Here's a picture of Skyler with Legolas, but you could also talk to Lora Croft, Buffy, or one of these outstanding problem solvers.

And if teddy bears are a little too cute for you, try one of these guys from what is my favorite toy store on the planet, Kid Robot. (Next time you're in San Francisco, be sure to stop by... right down the street from the incredible Amoeba Music store).

Speaking out loud activates parts of your brain not accessed when you simply think. And even better, the act of explaining something gets you thinking in ways you might never have reached if you hadn't tried to talk through it. We often tell students that there is usually NO better way to learn than if you try to help someone else understand the thing you're still working on. Learn-by-teaching is even better than just stopping with learn-by-doing. We encourage members on the javaranch forums to try to get involved and answer questions, even at the risk of being wrong, because it helps the question-answerer as much as the question-asker.

And may I make an additional suggestion on the rubberducking/teddy bear technique... it works especially well on pets. If you can get them to sit still long enough, you might be surprised how much they'll learn in the process.
My sheep (R.I.P.) Albert used to be one hell of a coder, and was especially good at brainstorming.