November 18, 2012 @ 1:49 am
[Note for CoE and direct evolution relevance: Towards the end of this HHMI Holiday lecture is a really good discussion about the difference between the inductive data-gathering + intuitive leaps method of Newton and Darwin and the deductive elimination-of-hypotheses method that I talk about below as "the" scientific method. Writing stuff down is equally important for both, and (I would say) for any other formalization of our trial-and-error learning behaviors.]
My teaching mentor Mac Frank used to make fun of me for time-stamping my teaching journal entries. 2 in the morning is a workaholic time, or a night-owl time, usually. In reality, I went to bed at 9pm and woke up to empty my bladder; then, as often happens to me, the ideas start up, and the quickest way to get back to bed is to go ahead and write them down – exorcise the muse, to mix a metaphor. Otherwise they just keep cycling in my head like verses from a song that is stuck on repeat, and I’ll be up for hours. Right now, it’s “Jumping Jack Flash.”
Yesterday the mistress of the VSI Facebook page posted a Mythbusters quote from Adam Savage, something like
The difference between screwing around and doing science is writing stuff down.
I added the word "doing" because it restores the rhythm of the line, poetically (at least I think so; I'm a really bad poet). I love that line.
Coincidentally, I spent all week with my Honors students working on that exact issue. There’s a little virtual laboratory setup called Frog Pond that I’ve been using for years now. It’s just rich enough to reveal to students that while they can memorize and repeat the steps of the scientific method, they don’t actually understand it yet. Applying the method in practice is much more difficult than it sounds.
Most of the time, screwing around in real life, working on some problem, we consciously or unconsciously have three or four hypotheses competing in our heads at the same time – only the most likely ones, based on our prior experience (like Mac’s socially based assumptions about my motivations for time-stamping things, when in fact I'm just half-assedly tracking my insomnia). That’s fine, most of the time, but what if you’re doing research? What if your patient’s disease is brand new, like this story about an autoimmune disorder that was only discovered in 2009? That’s where the rigor of the scientific method comes in especially handy, on the variation side of VSI, laying out not just the probable hypotheses but the possible hypotheses. The Periodic Table was brilliant because it laid out all the mathematical possibilities for new elements:
“Let’s see; we have one with 5 electrons in the outer shell – that’s nitrogen – and we have one with 7 – that’s fluorine – so there must be one with 6 that we haven’t found yet. Let’s look for that one!"
We could apply the same logic to medicine, saying, in essence:
"Here's all the proteins on the surface of our brain cells. Here's all the ones we know of that trigger autoimmune responses. Where are the holes, the proteins that we know exist but that we haven't yet found autoimmune symptoms for?"
Big job, but so was the Periodic Table, at the time.
With Frog Pond, the students have real difficulty stating anything new rather than just repeating what they already know, or think they know. They have difficulty generating all the possible hypotheses, instead of only the most likely based on prior experience, which are the first ones that bubble up into consciousness from their associative memory banks. They don’t write their hypotheses down, so that the ideas stay still on the page, instead of swirling around in their heads and getting confused. They have difficulty stating what they think the results of an experiment will be in enough detail to interpret those results, or to tell anyone how to replicate their experiments. Really, I should say “we,” because we all have those same problems. Otherwise we wouldn’t need to learn a scientific method. Screwing around would be good enough.
So because Frog Pond requires the students to design their own genetics experiments, and because screwing around wasn't good enough in that case, on Friday I stepped back and found this neat little problem set, which does exactly the process I've been trying to get the students to internalize. Every step for each problem is laid out in the sort of detail that professionals tend to do intuitively. At this stupidly explicit level of detail, the problems become easy. It's trying to get students to stop taking shortcuts that look like they will save time (but just confuse things) that is hard.
Postmodernists will tell you that there is no scientific method, that screwing around is all there is. In one sense that’s true; the Western Renaissance scientific method is not a law of the universe so much as it is a technique for getting at the laws of the universe, a formalized system that simplifies our natural screwing-around behaviors, losing some of the richness of those behaviors but condensing them into a memorable, communicable sound byte meme, something we can actually talk about. But that’s true of any system, any method, and science is a pretty damned good method, compared to all the others we’ve tried as a species. As an analogy, think of the different traditions of meditation, many of which predated the Buddha, some of which came after. A postmodernist is like someone who’s never meditated telling a committed, practicing Buddhist that meditation doesn’t work – not only that, that there’s really no such thing as meditation. Doing science is like doing meditation; you have to do it in order to understand it. Otherwise the fruits of those labors, like technology for science or personality changes for meditation, look like magic.