Wednesday, February 9, 2011

Related Work Considered Harmful?

"I can come up with an innovative way to do X because I don't know how X has been done in the past."

I hear this comment from time to time. The idea behind the comment is that somehow, not understanding the related work in an area, not having been "brain-washed" by existing ideas, is a key ingredient to innovation. The fallacy of this idea bothers me so much that I've come out my 3-year blog dormancy to write about it.
Let's first establish that this idea is at best anecdotally true. In fact, I'm being generous when I say it's anecdotally true, since I am not aware of any such anecdotes. I am, however, aware that Einstein did not develop the theory of relativity because he did not understand physics; Steve Jobs and Steve Wozniak did not build the personal computer because they didn't understand how mainframe computers worked; and Jim Lahey did not develop the no-knead bread because he didn't know how bread can be made with lots of kneading.

In fact, all of these above people understood the related work in their respective areas extremely well. So well that they could pinpoint exactly what is inefficient or incorrect about the state-of-the-art, and thus develop new theories and techniques to address those inefficiencies or incorrectness. That is the difference between an expert, and an amateur. An expert can tell you exactly why he made the decisions he made, how it compares to all the other alternatives. An amateur, on the other hand, says something to the effect of, "it felt right".

The true test of an expert as well as an innovator is to be able to understand his related work so well, yet still be able to remain critical. He is able to see through the smoke and covers surrounding the related works, understand the essential ideas, yet still critically examine these ideas and how they relate to what he's trying to accomplish. Only then does true innovation arise.

Has innovation ever been done by someone who didn't know his related work well? Perhaps. But such occurrences are probably exceptions rather than the rule.

Let's conclude with the following anecdotes: Tim Berners-Lee did come up with the URL/HTTP/HTML specification with passing knowledge of parsing theory; and Will Rogers did come up with the "trickle down economics" because he's not an economist.