Saturday, October 12, 2013

A carefully planned sneeze

I'm working on job talks for tenure-track Assistant Professor positions at schools around the country. Crafting these talks has provided me with the insight that doing experimental science in academia is a lot like having the urge to sneeze, sneezing, then discussing the sneeze in detail as if that particular sneeze were the logical and only thing to do at the exact time you did it.

Each talk I'm writing combines a discussion of the motivations for each of 2-3 experiments together with a discussion of the results and their implications. It seems like it would be easy to write the talks in the proper order -- my motivation always occurred before the experiments and results in real time, right?

It depends, I think, on whether we're talking about the real motivation or the stated motivation.

The real motivation that I have for doing an experiment is essentially always the same. It's something like, "Wouldn't it be cool if blah blah blah influenced blah blah blah and the mechanism was blah blah blah? Yummy!" The thought comes out of thin air, seemingly, and it's not attached to anything rational or even a memory of the reading that might have influenced it. It's exactly like having to sneeze. It comes over me without effort and sometimes at awkward times.

What I am starting to believe is that this is true for most scientists, but apparently because we want to be perceived as rational beings (good luck, us!) we have agreed that it's best not to talk about the shameful secret of our irrational experimental intuition. Instead we sort of retrofit rational explanations to our motivations. "Goldberger and Schnitz showed in 1989 that blah blah blah, so I thought blah blah blah -- which was clearly a reasonable single step from this previous idea."

I'm not saying that Goldberger and Schnitz' seminal work (ovular work? does it depend on their gender?) is not important, nor that scientists shouldn't cite related work, nor that Goldberger and Schnitz couldn't have influenced my thinking subconsciously.

However, if what happens to me is similar to what happens to other scientists, then the conscious order of events in the world of experimental science is more like this:

1) Get an idea for an experiment, probably while showering.
2) Look up some references and read/re-read the ones related to the idea.
3) If the idea is still feeling like a strong urge, ignore any references that go against the idea.
3.5) ...even if there many many references that go against your idea, and very very few (or none) that support it.
4) Design and perform the experiment.
5) Look at the data and try to understand what they are telling you.
6) Go back to re-reading references in order to understand why you should have predicted the result that you got.
7) Write a story that makes you appear rational throughout this process.

It's somewhat embarrassing to write this process down, even though I am sure that I and many of my colleagues do exactly this. We're trained not to report our intuitive urges, but we all have them. Once we do have them, and if we find them valuable, we have to come up with a story about how smart we were to have had them.

Sure, it's the same rationalization process everyone uses, including non-scientists. It's just that we scientists are supposed to be empirically observing and recording what is actually happening.

It just seems a bit dishonest to carry on as if we didn't simply have to sneeze.