I was listening to a podcast where Tyler Cowen interviewed Ed Boyden (linked here), a fascinating professor at MIT. Boyden presents the “tiling-tree” problem-solving technique, which is about thinking backwards from a problem and enumerating through all possible tools for solving the problem. I thought it was a really fascinating and systematic research approach.
Boyden gives the example of optogenetics (his research field), where he framed the problem as: how can you control the brain to repair it? The group began thinking about how you could control the brain and realized there are only so many different kinds of energy: mechanical force, magnetism, electricity, light. Then from each of these ideas, he would break them down further into discrete “nodes,” and then he would test the bottom nodes with a literature search or doing an experiment.
Apparently, Fritz Zwicky was the pioneer of this thinking (Zwicky called it “morphological analysis”) in the 1930’s. He used this method successfully to imagine (and then discover!) dark matter and a bunch of other ideas in astrophysics.
In essence, Boyden is reasoning from first principles in this problem-solving approach.