Thursday, December 14, 2006

Wise fools

A few days after writing this post, I've found some great short papers: How To Be a Good Graduate Student, How To Do Research In the MIT AI Lab, and an old favorite of mine, a speech by Hamming (yes, that Hamming) entitled You and Your Research. Grad school, how I look forward to thee. I'm getting ready, and I'll be there someday.

In other news, Chandra, Eric Munsing, and Jon Tse have all confirmed (over pizza) the existence of the following behavior pattern.

The Wise Fool Phenomenon: The probability that you will ask unashamed intelligent questions about a topic is inversely proportional to the amount you believe you are supposed to know about that topic.

For instance, I'm much more likely to ask about mechanical engineering topics unashamedly, without fear of "looking stupid," than I am about electrical engineering ones. Ironically, this means I learn about mechanical engineering more rapidly when I decide to learn something about it. I recognize this is a stupid thing to do, and I'm trying to change it. It's one of the (many) reasons I don't do well in classes.

The best explanation Chandra and I could come up for this is that you think you're already "supposed to have learned it," and that you are therefore being stupid (and will appear as such) and wasting everyone's time by having them explain things you should already know. If you aren't "supposed" to know it, your questioning (for some reason) provides amusement/insight/warm fuzzy feelings for the people you're asking questions of... and besides they can always say no, because you don't "have to know it" anyway. When it seems like you should already have something in your cup, it's very hard to empty it.

Corollary: You are more likely to exceed expectations when you don't know what they are. This assumes you've got an initial interest/aptitude in the subject and are pursuing it. If you're told to go to height H, you tend to go to H (or a bit higher if you want to look especially impressive), and then stop because you've "succeeded." If you can't make it to H, you "fail." On the other hand, If you aren't told a set height to reach, you just climb, and as long as you like it, you just... keep going.

There is a certain height at which it's reassuring to have someone look up and exclaim in amazement at how far you've gone, but usually by this point you're well past any H they would ever have set. If you don't make it far, that's okay; you haven't "failed," just chosen not to continue. Besides, at some point, the folks who advance any field must tread places nobody has ever gone before; why not start the process of trusting your own learning earlier?

There's a flip side to both of these as well. An initial goal or requirement can help you discover that something's there to learn, and give you a starting point at which to look. Lack of useful metrics makes it more difficult to reflect on your own learning. One reasonable balance is the idea of minimum and maximum deliverables (as Allen Downey calls them; Rob Martello calls them "circles") where you set an easily-achievable bar as an absolute goal but also toss out, as a tentative target, the bluest-sky dream you could ever imagine reaching. That way you get something done, but have freedom to do more (and a vague starting point to head towards if you haven't found something more agreeable by that time).

But that's getting off track. As far as I can tell, you can take advantage of Wise Fool Phenomenon by:
  1. Learning things "before you're supposed to know them" (the reason I used to do ridiculously well in math classes; I'd already read books and asked lots of stupid questions about the stuff before we got to it in class).
  2. Beginning a learning endeavor by making the big disclaimer that you know nothing and will be asking lots of stupid questions.
  3. Actually asking lots of "stupid questions."

It's the last one that gets me. I don't feel like I've got the right to ask "stupid questions" unless I've been working hard and doing my utmost to keep up - I feel like I've got to go as far as possible by myself before calling for help. This works really well in the cases where I actually take the time to go as far as possible by myself before calling for help.

Unfortunately, "as far as possible by myself" is pretty far, so I usually don't get there. This means I often consider myself to not have "gone far enough." When I fall behind in classes this leads to "but I can't ask for help because I haven't worked hard enough" syndrome, which leads to me putting off talking to my prof until I've caught up on my own to "prove myself worthy," but I can't catch up right away because I'm already behind, which leads to me putting off asking for help even more...

Lynn says I need to be less afraid of wasting people's time, and the very fact that I'm afraid of it means that I typically don't do it. To that, I'd add that being afraid of wasting their time in the beginning can also lead to wasting more of their time at the end when I mess up because I didn't ask first, so ignoring the Wise Fool mandate is just a really stupid thing for me to do.

This has been Exhibit A in the "Why Mel Isn't Ready For Grad School Yet" series.

2 comments:

Anonymous said...

There are two situations in which I will ask questions:
1) When I'm not supposed to know the answer and also personally feel no shame that I don't know it. This is pretty much limited to non-engineering stuff.
2) When I have spent so many hours on a particular thing that the probability that my question is actually stupid is very, very low.

/me shrugs

Becky

Mel said...

Thanks, Becky - that does about sum it up.

Regarding #1: One of the initial reasons I went into electrical engineering (besides the fact that the dart happened to hit that section when I chose my major) was that it was the thing I was most terrified of. Weirdly enough, I might be less afraid of asking questions about it if I wasn't an ECE. On the other hand, I wouldn't have been forced to develop this confidence (or at least some semblance of a front of confidence) that causes me to experiment with the material, had I not declared that I was an ECE and therefore Should Know Stuff Dammit.

Regarding #2: I always feel kind of weird for not asking earlier once