Science is not math

I probably should have included this idea on my "All about science" blog post. Maybe I didn't put it in there because if I talk about what science is you can figure out what it is not.

Science is not math

Science is all about making models. It is true that many current models are mathematical models, but it doesn't have to be that way, and it hasn't always been that way. Think about rubbing a piece of metal with a magnet. It gets magnetized - right? What if you then cut that magnet in half? Then you have two smaller magnets. How can you make a model that explains this phenomena? Yes, you could use some mathematical models, or you could draw a domain model like this:

i-d3626449b1232bd89a21efdd46eba9d3-domain.jpg

No math needed. Or what about biology? I know there is a lot of biology that uses some math, but not all of it - right?

Also, you could say the following.

Science is communication

I was going to say "english" but clearly that is not true. Can you do science without communicating? I guess technically, yes. But every example you would look at would have people writing stuff down and communicating with other people.

So, in summary, if you want to say one thing that science is, I would say Science is the building and testing of models.

Oh, I guess I should mention this xkcd comic. It shows fields arranged by purity. It does not say that math is a science, it says it is the most pure (field).

More like this

Mathematics is self-consistent and survives abstract proof. Science is self-consistent plus survives empirical falsification. Everything else interpolates, curve-fits, or simply babbles, then screams "heteroskedasticity!" or "test of faith!".

Things stink when a model offers limitless untestable outputs (string theory) or it is buttered toast with jam yesterday and tomorrow but only dry stale bread today (religion). Things stink when exotic untestable extrapolations (Standard Model extensions) crowd out testable alternatives (Equivalence Principle parity violation by chemically identical, opposite parity atomic mass distributions - crystals in space groups P3(1)21/P3(2)21 or P3(1)/P3(2)).

Texas recently declared by committee majority vote that the entire universe is a biblical 6000 years old. One presumes Andromeda's ballot was rejected along with those of cosmic background radiation (fluctuations); mica decay halos, C-14 in any oil well or coal seam, Oklo, Gabon; tholeiite basalts of the New Jersey Palisades and Staffa, Scotland; magnetic banding of the Pacific Ocean floor. Tests of faith.

I think science is the act of trying to understand, using logical arguments and experiments, isn't it? All the other stuff like faith are not science. The untesteable outputs (like string theory) are just rational hypothesis (well, sometimes "almost-rational" hypothesis). When a scientist has an idea of how test it, then becomes science.