Science and Predictability

What is science? Thanks to the insane accuracy of physics, many people believe that that science means the ability to predict. But is that the right definition, wonders Mitchell Waldrop in Complexity. Or is the ability to explain things the correct measure of a science?

 

He cites several good examples in that context.

“Was Darwin “unscientific” because he couldn’t predict what species will evolve in the next million years? Are geologists unscientific because they can’t predict precisely when the next earthquake will come, or where the next mountain will rise? Are astronomers unscientific because they can’t predict precisely where the next star will be born?”

 

Then Waldrop cites everybody’s favorite problem field – economics. When the market and the economy is unpredictable, why do economists insist on making it like physics? Why come up precise mathematical equations, built on the assumption that all humans are perfectly rational (and thus predictable), when (a) humans are so obviously irrational, (b) markets fluctuate wildly, and (c) economic cycles are so unpredictable?

 

If chaos theory with its unpredictability is accepted as physics, shouldn’t other fields that are not predictable be willing to be like chaos theory – science without the predictability?

 

Harold Agnew makes a similar point, that the physical sciences are full of “conceptual elegance and analytical simplicity”. But so many other fields, including molecular biology, don’t fall in that category:

“Once you’re in a partnership with biology, you give up the elegance, you give up that simplicity. You’re messy.”

This approach is necessary in the social sciences too – they’re messy. But “messy” can still have some deep principles underlying them; it’s just that the interactions are so many and so intricate that it makes the final outcome unpredictable.