What is science?

I am thinking about the nature of science on at least two levels. For one thing, we are trying to learn what the teachers in our program think is scientific, because one of our goals is to have them recognize science to be what we think of it as being. For another, I am thinking about the kind of research that I do these days and trying to identify what makes it scientific (and when it falls short). 

Here are a couple things that are not necessarily defining characteristics of science for me:

1. The scientific method. Though hypothesis testing and control of variables is great stuff, it can't be the whole story. At a minimum, there has to be some prior activity by which you obtained the hypothesis.

2. Laws of nature, by which I mean F=ma, energy conservation, and other such things. At a minimum, there has to be a process by which such laws are developed, and that would mean being scientific even before you knew anything about any laws.

I think that either of these things might be sufficient for identifying something as scientific, depending on your values, but to me neither of them is necessary - there is science that has neither of these properties. Hunter introduced me to Robert Hooke's drawing of a flea:

I mean, look at that thing. Click to make the picture bigger -- it's incredible. I had no idea a flea looked like that. Though there is no hypothesis here, no claim, and no laws of nature, to produce such a drawing feels like high science to me. So, here is something I count as science: The systematic observation and documentation of the details of phenomena.

This can't mean just looking at things. The explorers who went down into the deep ocean in a bathysphere, Barton and Beebe, likely saw all kinds of astonishing things, but didn't bring back much that oceanographers felt they could use, partly because they were unable to document what they saw and partly because they didn't know all that much about fish. Their seeing was limited by their lack of professional vision (the distinctive ways in which professionals notice and interpret phenomena of their profession). So, whatever it is that makes observation scientific has to do with having a "trained eye." And the observations that are made have to answer to the standards of the discipline.

Another thing that I want to count as science is going to be something about model-building -- the construction of explanatory accounts of phenomena. I haven't thought as much about that today.