9 January 2019 – This week I start a new part-time position on the faculty at Florida Gulf Coast University teaching two sections of General Physics laboratory. In preparation, I dusted off a posting to this blog from last Summer that details my take on the scientific method, which I re-edited to present to my students. I thought readers of this blog might profit by my posting the edited version. The original posting contrasted the scientific method of getting at the truth with the method used in the legal profession. Since I’ve been banging on about astrophysics and climate science, specifically, I thought it would be helpful to zero in again on how scientists figure out what’s really going on in the world at large. How do we know what we think we know?
While high-school curricula like to teach the scientific method as a simple step-by-step program, the reality is very much more complicated. The version they teach you in high school is a procedure consisting of five to seven steps, which pretty much look like this:
- Observation
- Hypothesis
- Prediction
- Experimentation
- Analysis
- Repeat
I’ll start by explaining how this program is supposed to work, then look at why it doesn’t actually work that way. It has to do with why the concept is so fuzzy that it’s not really clear how many steps should be included.
The Stepwise Program
It all starts with observation of things that go on in the World.
Newton’s law of universal gravitation started with the observation that when left on their own, most things fall down. That’s Newton’s falling-apple observation. Generally, the observation is so common that it takes a genius to ask the question: “why?”
Once you ask the question “why,” the next thing that happens is that your so-called genius comes up with some cockamamie explanation, called an “hypothesis.” In fact, there are usually several possible explanations that vary from the erudite to the thoroughly bizarre.
Through bitter experience, scientists have come to realize that no hypothesis is too whacko to be considered. It’s often the most outlandish hypothesis that proves to be right!
For example, the ancients tended to think in terms of objects somehow “wanting” to go downward as the least weird of explanations for gravity. The idea came from animism, which was the not-too-bizarre (to the ancients) idea that natural objects each have their own spirits, which animate their behavior: Rocks are hard because their spirits resist being broken; They fall down when released because their spirits somehow like down better than up.
What we now consider the most-correctest explanation (that we live in a four-dimensional space-time contiuum that is warped by concentrations of matter-energy so that objects follow paths that tend to converge with each other) wouldn’t have made any sense at all to the ancients. It, in fact, doesn’t make a whole lot of sense to anybody who hasn’t spent years immersing themselves in the subject of Einstein’s General Theory of Relativity.
Scientists then take all the hypotheses available, and use them to make predictions as to what happens next if you set up certain relevant situations, called “experiments.” An hypothesis works if its predictions match up with what Mommy Nature produces for results from the experiments.
Scientists then do tons of experiments testing different predictions of the hypotheses, then compare (the analysis step) the results, and eventually develop a warm, fuzzy feeling that one hypothesis does a better job of predicting what Mommy Nature does than do the others.
It’s important to remember that no scientist worth his or her salt believes that the currently accepted hypothesis is actually in any absolute sense “correct.” It’s just the best explanation among the ones we have on hand now.
That’s why the last step is to repeat the entire process ad nauseam.
While this long, drawn out process does manage to cover the main features of the scientific method, it fails in one important respect: it doesn’t boil the method down to its essentials.
Not boiling the method down to its essentials forces one to deal with all kinds of exceptions created by the extraneous, non-essential bits. There end up being more exceptions than rules. For example, the science-pedagogy website Science Buddies ends up throwing its hands in the air by saying: “In fact, there are probably as many versions of the scientific method as there are scientists!”
A More Holistic Approach
The much simpler explanation I’ve used for years to teach college students about the scientific method follows the diagram above. The pattern is quite simple, with only four components. It starts by setting up a set of initial conditions, and following two complementary paths through to the resultant results.
There are two ways to get from the initial conditions to the results. The first is to just set the whole thing up, and let Mommy Nature do her thing. The second is to think through your hypothesis (the model) to predict what it says Mommy Nature will come up with. If they match, you count your hypothesis as a success. If not, it’s wrong.
If you do that a bazillion times in a bazillion different ways, a really successful hypothesis (like General Relativity) will turn out right pretty much all of the time.
Generally, if you’ve got a really good hypothesis but your experiment doesn’t work out right, you’ve screwed up somewhere. That means what you actually set up as the initial conditions wasn’t what you thought you were setting up. So, Mommy Nature (who’s always right) doesn’t give you the result you thought you should get.
For example, I was once (at a University other than this one) asked to mentor another faculty member who was having trouble building an experiment to demonstrate what he thought was an exception to Newton’s Second Law of Motion. It was based on a classic experiment called “Atwood’s Machine.” He couldn’t get the machine to give the results he was convinced he should get.
I immediately recognized that he’d made a common mistake novice physicists often make. I tried to explain it to him, but he refused to believe me. Then, I left the room.
I walked away because, despite his conviction, Mommy Nature wasn’t going to do what he expected her to. He persisted in believing that there was something wrong with his experimental apparatus. It was his hypothesis, instead.
Anyway, the way this method works is that you look for patterns in what Mommy Nature does. Your hypothesis is just a description of some part of Mommy Nature’s pattern. Scientific geniuses are folks who are really, really good at recognizing the patterns Mommy Nature uses.
If your scientific hypothesis is wrong (meaning it gives wrong results), “So, What?”
Most scientific hypotheses are wrong! They’re supposed to be wrong most of the time.
Finding that some hypothesis is wrong is no big deal. It just means it was a dumb idea, and you don’t have to bother thinking about that dumb idea anymore.
Alien abductions get relegated to entertainment for the entertainment starved. Real scientists can go on to think about something else, like the kinds of conditions leading to development of living organisms and why we don’t see alien visitors walking down Fifth Avenue.
(FYI: the current leading hypothesis is that the distances from there to here are so vast that anybody smart enough to figure out how to make the trip has better things to do.)
For scientists “Gee, it looks like … ” is usually as good as it gets!