What is Science? A school kid’s definition goes something like this: Find a hypothesis (from somewhere); make sure it is falsifiable; test it against reality; if it fails, discard it; if it doesn’t, published it. Rinse and repeat. We’ll call this SKD view of science for shorthand.
There is some truth in it. In the same way that, being a good tennis player means, being able to hit the ball really hard, keeping your knees bent, and keeping your eye on the ball. While that’s got some things right and that seem to lean somewhat in the direction of what it means to be a good tennis player, there is much that could be taken away and gobs of stuff that could be added to give a richer and more accurate description of the concept.
The SKD is used as a weapon by some who want to lend their ideas the aura of science, and by ‘aura,’ I mean the credentialing and authority that comes with the label science, without understanding how science works or how it is practiced. Or why it is so powerful in explicating the world. I’m thinking in particular of the anti-evolution, and the climate change denier, crowd. Some of these folks’ tactical approach is to discredit the work of scientists because they fail in some way the SKD view and therefore is not ‘true’ science (Note: whenever someone puts ‘true’ in front of science, let it serve as a warning that you are about to be introduced to something that is ‘not’ science.)
So what is science? First let me explain what it is not. Science’s (and I’m going to act as if science is a ‘person,’ or a unified concept in this post, and say things like ‘Science is’ and ‘Science does not permit x’ as a handle or shortcut, but as you’ll see I will subvert this use later and show that it is neither unified nor a single monolithic thing, but bear with me, it does mean something), anyway science’s, upfront, nonnegotiable stance is methodological materialism. This means no hidden forces. No influence from God, angels or demons. No magic. No miracles. This does not mean that scientists don’t believe in God, or miracles, or that science claims that nothing that does not fit its materialist claims is worth knowing. No, science does not claim that it will reveal all truth, in fact it really can get little purchase on lots of things we make value claims about like art, ethics, religion, etc. It doesn’t even claim to get at things (this despite misguided attempts by the likes of Richard Dawkins to claim that it can discover all truth). So science is not a method that speaks to all truths of every kind.
Some seem to be afraid of methodological materialism. But you are very familiar with methodological materialism. It’s what you expect from your car mechanic. She assumes that whatever is causing the clunking noise in your DeSoto, it is a mechanical problem, with a particular cause, and that she can take actions to correct it based upon new car parts, tightening, or loosing, bolts, or some such action based purely on the physical realities about the ways cars are put together from metalish things, lubricants, gadgets, and such. If she said, “It looks like malicious fairies have given the engine a curse that causes dark fluxuals from the netherial world of Kandoonianus.” You would likely get a new mechanic. Not that there might not be a curse from said fairies, but that’s not the way to bet, and you expect, and your experience with the world suggests, that the best way to approach car repair is from the perspective of methodological materialism. This assumption is science’s best move too. For exactly the same sorts of reasons. Your mechanic may be an atheist, Buddhist, or Mormon, but this is irrelevant to how she investigates your car. She assumes it is a mechanical problem and moves from there, regardless of what spiritual commitments she might have.
Methodological materialism =/= no threat to spirituality. Methodological materialism = good scientific assumption.
So defining science:
Here’s a nice practical definition: Science does the best things it can to explicate the world using a bunch of tools that have worked so far. What? Yeah. Sort of a minimalist description, agreed, but let’s roll with it for a minute and unpack that word ‘tools.’ Or rather unpack what some of those tools are. In addition to the tools, let’s look at stances or postures and attitudes that science takes.
Experimentation: Yes. Hold as much as you can constant. Simplify the world as much as you can. Then manipulate something and if you have controlled everything else then the relationship between effect and your cause must be, well, causal. A great deal of effort as been worked out in getting these experimental tools down. Taking measurements, handling the data, randomizing things so your own biases don’t get in the way and you can average out other effects you aren’t interested in, statistical analysis, great stuff. So experimentation has been important in science. It turns out easier to falsify rather than confirm things, so you usually want to aim for that. Hence the KDS above is not a bad abstraction/simplification of some really hard stuff (see Fleck #1 below).
Observation: Well as much as we would like to do experiments, the world is too big and complex to pull it off all the time. Take astronomy. Getting galaxies into the beaker has proven fairly hard. Ecological systems too. Geology too. Emergent behavior—very tricksy. When the parts do not sum to the whole, and the whole influences the parts, everything you want to do with science gets trickyery. You still want to measure and stuff. Correlate. Hypothesize on what you think will happen if other stuff happens. You want to carefully gather data, categorize, systematize, organize, any –izing you can do, you should. You want to propose explanations of what you see. All very important in science. No SKDing here though. Or very little anyway.
Modeling: If you believe in a world that is causal you ought to be able to mathematicize things. Math is a formal way of writing down quantitative descriptions of how some things are thought to influence other things. Sometimes a good causal theory can be written down with a few strokes of a pen—like Einstein did. Then using that math, predict things, if you are good at predicating it’s a good sign you’ve got a good model. What more could you ask for from a model? Things have gotten more complicated with computers, and the range of phenomena you can describe has skyrocketed. Most important, modeling brings together experimentation and observation in helpful ways. These three things constitute the main tools of science. Models of course are abstractions of idealized worlds and should never be taken for reality, but they are very useful for testing whether you understand how the world works. They have both explanatory and predictive power when they work right, which are two of science’s highest values.
So given the tools. What activates define and constrain science that make it so darn powerful for explaining the world?
A stance of openness to revision and holding results as tentative: Science is very humble. It has to change its mind sometimes. Conclusions are tentative. New facts, new analysis, new interpretations, sometimes force a confrontation with old facts, old analysis, and old interpretations. Science thrives on this. It holds as open all its findings. Not to change them willy-nilly, mind you. No, science is more than a list of suggested ideas to hold onto this week. Its claims have been but into the furnace for testing, heated, then sledge hammered to see if we can get the claims to crack. When they don’t we gain confidence we are onto something. We always know we might find a hotter furnace or a bigger hammer, so science is ready to change. But it does not bow to claims that it might be melted or cracked—you actually have to do it. For example, Intelligent Design creationism has been waving hammers in the air for fifteen years shouting, ‘we could crack evolution anytime we want.’ Except they haven’t actually given it a whack yet.
Peer Review and Publication: Science is not the Internet. All voices don’t get a say. Your voice has to pass muster. Your claims have to be examined. Your analysis, your interpretation has to be scrutinized by experts. It has to fit into the context of other work that has been done. This is a bloody process. Science is a crowded field and only the best, most well tested, ideas get through the gauntlet. Then when something is published. It is still open to the scientific community for further scrutiny. The claims continue to be prodded, attacked, poked, repeated and replicated, and bothered until it gives, or people start to think there is something here worth looking at. If it isn’t published in the peer review literature, it’s not science. Hence the power of things like Anthropogenic Climate Change in which the peer review literature is united and the air-hammers of the internet say something else. Peer Review is vital to good science. If it is not playing the science game, it’s not science.
Research Programs: In science, it is within disciplines and their own research programs that best practices are established, through trial and error, traditions, lab experience, repeatability, instrumentation, training, agreement, finding what works and what doesn’t. New technologies, come and go, their use is tried against former methods and winners and losers are arranged. Collaborations are formed, dissolved. Students are trained and credentialed in these programs and improve ideas, challenge old ones, bringing their creative genius into the work. Changing and improving things. Not in some absolute way, but in practical ways.
So that is science. Not the clean ‘method’ that often get’s cartooned as what science is in the SKD. In short, it is a Darwinian process of ideas. Ideas, theories, hypotheses, are thrown into a struggle for existence. Fitness is defined as how well the claims confront the world and its processes. Only the best survive and to go on to reproduce. If something new comes along, it has to fight in the arena. Prove its mettle. Enter the gladiatorial contest and survive to fight another day. Science is practiced by people. People with all the same weaknesses, shortcomings, misalignments, as any group. But the structure is in place to create a dynamic marketplace of ideas. And nothing has come as close to explaining so much. It works and that is its highest recommendation.
#1. In his book, Fleck details how his research team discovered the Wasserman reaction for syphilis diagnosis. As dry as that sounds, it reads like a detective novel. Its import however lies in a careful view of how science works. Not in the observe-world—make-testable-falsifiable-hypothesis—test hypothesis—publish-results way, that he acknowledges is how things get written up, but in the nitty-gritty world of how science really works. He details how the research was filled with false starts, flashes of creativity and speculation, back tracking, back to the drawingboarding, team work, collaboration, trial and error, abandoned trajectories, instrumentation problems, how conceptual foundations had to be rethought, how he had to bring in knowledge and background education, the use of inference, induction and induction, and the ugly messiness that goes into real science. In fact he quips, “If a research experiment were well defined, it would be altogether unnecessary to perform it.”
Fleck, Ludwik, 1979. Genesis and Development of a Scientific Fact University of Chicago Press, Chicago.