Re:The 40,000-Mile Volcano
by HCIdivision17 on Tuesday January 12, 2016 @02:52PM
(#10889868)
Nicely put. This how I always think about economics: we've got plenty of numbers and models, but meaningfully large experiments are are to design without risking a lot of livelihoods. Science is brutal at scale, and I'm ever impressed with the techniques to make sense of the noise.
The natural sciences are incredibly complex. You're not dealing with nice simple interactions, and we can't run experiments on the time, spatial, or energy scales needed to directly test hypotheses in the way you would in an experimental science.
Instead, we observe.
We try to understand the present to understand the past and the past to understand the present.
When you're observing an uncontrolled system, it's very tough to isolate one variable from another. Geology is often accused of not being "quantitative enough". What people don't realize is that the quantitative part is usually stupidly easy. It's choosing the right physics to model that's hard. We usually don't fully understand exactly which physical process are at work.