What is Science?

When I was in grade school, our class was taught the Scientific Method. This method, which we were told guides all scientific research, involved the following steps:
  1. Observe some phenomenon in nature.
  2. Form a hypothesis to explain or model the phenomenon.

  3. Design an experiment to test the hypothesis.

  4. Carry out the experiment, recording all input data and output measurements.

  5. Analyze the results of the experiment.

  6. Determine whether or not the results support the hypothesis.

Suppose you are interested in determining the population of ancient Troy from its founding through it's abandonment. You consult historical records. You examine the ruins of the city, to determine its extent. Perhaps you conduct some experiments, carbon dating say, to determine the age of certain artifacts. From this collected data you make an estimate of the city's growth and decay. Is this science? In my opinion, it isn't. It's Quantitative History, but it isn't science. There is no experiment being made that can be verified with a new collected data set. There is no natural phenomenon that is being modeled and explained. While some of the steps in the Scientific Method fit to a certain extent to the process described, overall it doesn't really fit the working definition of science.

It is said that the "science is settled" about Global Warming. That the Earth has warmed 0.6C in the past century is practically written in stone. But was this conclusion based on science? In fact, it was based on the statistical analysis of temperature data taken at weather stations and ships at sea since 1860. It was not based on laws of physics from the theory of heat transfer and fluid dynamics at all. There was no repeatable experiment conducted. There is no ability to predict the future from the data; it is not even tied to an increase of greenhouse gasses. Like the analysis of Troy's population, this is quantitative history, not science.

The Earth may indeed have warmed over the past century. But when I hear that the "science is settled," I think to myself "what science?"


Anonymous said...

(your sis again.)

Most astronomy is also untestable and can rely on no true experiments, only predictions of what data will show.

Is astronomy, then, a science?

SteveBrooklineMA said...

I disagree about "most astronomy," but understand the point. Cosmology is often criticized in this way, and I think there is some validity in the criticism. I see a loophole: you can think of something like string theory, for example, as being in the hypotheis-forming stage of the scientific process.

SteveBrooklineMA said...

See for example David Lindley's
The End of Physics for some of that critisism.