Defenestrating Thoughts from the Bivouac

Back on the topic of framing… I was talking to a college professor about students and the state of education. I mentioned that I feel we don’t teach people to look at the larger picture and think through problems from the full end to end. For example, take corn-based ethanol, it seems like a smart idea to make gas out of corn. After all, corn is cheap and gas is expensive; may as well turn something cheap into something expensive. :-)

Now though, the price of tortillas in Mexico has almost quadrupled. And the price of meat is on the rise. It seems obvious now; if you drive up demand of something really fast, you’re going to raise the price and that will be reflected in the price of other things made out of corn. But, why didn’t we think of that at the start? Probably because we didn’t think through the problem and instead went for a myopic analysis.

The professor said that this is part of larger problem of not asking the right questions. Or more precisely, defining questions with a artificially limited scope. For the issue of ethanol, the question was, “Which crops should we make ethanol out of?” The question should have been more along the lines of, “Should we make ethanol out of crops?”

I was watching the Today Show yesterday morning and one of the stories was, “Is the government doing enough to protect our kids?” This framing was interesting because it assumes that it is the government’s job to protect our kids. A more valid question is, “Who’s job is it to protect our kids?”

A better example is the “surge” in the Iraq War. For weeks all the analysts were debating, “Should the surge be 20,000 or 50,000?” or “Should the surge last 2 months or 6 months?” These questions are limited in scope; they are tactical questions about a strategy. In other words, they assume the strategy (the “surge”) is the correct course of action and limit the scope of the conversation.

A better question would be, “What are our goals in Iraq?” and then ask, “How do we best accomplish them?” Once you have a good answer for this, then you start tackling the tactical concerns like troop size and deployment schedule.

To put this in terms that geeks can understand, if you start a project debating, “C# or C++”, your project is doomed. First figure out what your project is trying to accomplish, then figure out which language is best for it. Next time someone asks you whether you prefer C# or C++, ask them what you’re using the language for. Without the right context, that’s like asking, “Do you prefer a spoon or a hammer?” :-)