Science has been one of my main interests since I was in second grade, and over the years I have had at least a passing interest in many different branches of science. Recently, I have tried to think about what I consider to be the overarching principle of science. What separates science from non-science?
There has been much written about this, some of it here. There are many different definitions of science, with some debate amongst different proponents. Karl Popper and Ludwig Wittgenstein (supposedly) almost came to blows over the issues, and while the prospect of the two fighting with a poker over logical positivism might be entertaining, the dense terminology used to debate induction, deduction, logic and utilitarianism, and many of the other buzzwords that surround the philosophy of science, are something of interest only to a few. And, if I can make a generalization, they are perhaps not something that working scientists are that concerned with.
Coincidentally, however, it was Wittgenstein that provided one of the best analogies for how to describe science. In Philosophical Investigations, Wittgenstein uses the example of the word "game", and its many contexts. Solitaire, tennis, jump rope, Candy Land, tag, golf, Sudoku and water polo are all games. Some games are competitive, some games are not. Some games are played with others, some are played alone. Some games involve physical activity, some do not. Some games involve luck, some involve skill, and some involve both. Further examples could be given, but the point is fairly obvious: there is no single definition that precisely fits everything that might be a "game".
And I have finally decided that science is the same way. There is no single definition that covers everything that might be described as "science". Some science is experimental. Some science is observational. Some science involves things that can be handled and built, and is therefore of utilitarian interest. Some science is very speculative. Some science involves almost solely abstract mathematical models, while some science just involves noticing obvious daily phenomenon. Some science deals with universal laws, while other science describes unique items or events. For example, it is a scientific observation that I saw a group of Stellar's Jays last week, in the Sapphire Mountains. It is scientific speculation to imagine that planets surrounding stars in the Magellanic Clouds would be mostly made of silicon and oxygen, with relatively small iron cores. And it is scientific calculation to compute the force of the gravitational attraction between two neutrinos a light year apart. In the first case, I am observing a unique event, and am not formulating a hypothesis. In the second, I am using known facts to guess at unknown facts, which I have no way of confirming or falsifying. And in the third, I am using mathematical laws to calculate a force that is too small ever to be observed directly. All three of these examples are parts of scientific activity, and yet they have little in common with each other, and no single overarching principle.
In other words, science is not a single, definable activity or discipline, but rather a term we used for a group of interrelated activities. While it may seem like a poor answer, I do believe, to paraphrase a supreme court justice, that the only real definition of science is "I know it when I see it".