Falsifiability is the ability of a theory to be proven false. Scientists talk about falsifiability as a criteria by which to consider a scientific theory. A scientific theory that is not falsifiable should not be considered seriously.
For example, suppose I had a theory of that gravity is caused by an invisible witch, who, every Sunday, casts a spell on the universe to keep gravity working for another week. Now, the witch lives on the dark die of the moon, she’s invisible, and she cannot be detected by any conceivable instrument (e.g. heat sensors).
Well, honestly, for all we know, this could be true. We don’t have a very good idea of what makes gravity work. But we still aren’t very interested in seriously considering this theory. Why? well, besides the fact that it involves a witch (or a fairy, if you have young children), it’s impossible to prove it wrong. Scientists are only interested in theories that can be proven wrong.
Now, this may make no sense. Why are scientists only interested in theories that can be proven wrong? Shouldn’t they be interested in theories that can be proven right?
Well, sure, but the problem with theories that can’t be proven wrong is that they might be wrong, but we would never be able to prove it. And that’s very troubling. Consider the Witch Theory of Gravity (WTG). It is (at least I think it is) wrong, but there is no way in the world I can prove it wrong. You can even find the Higgs Boson, but I would just say that the Higgs Boson is a result of the witch’s spell (on Sunday, probably while we’re all so innocently attending the evening service).
Such a theory is utterly uninteresting and unhelpful for the scientist. It’s not even worth investigating, regardless of its truth.
On a sidenote, this is one of the issues that the new atheism has with theism. Dawkins, for example, argues that theism is unfalsifiable—there’s no way to prove it wrong. And that is one point against it. (Dawkins also seems to think that he’s implicitly proven it wrong, but that’s beside the point.)
One area where I think scientists completely mess up their thinking about falsifiablity and unfalsifiability is in the debate over macro-evolution. Scientists often claim that creationism is unfalsifiable. It posits an unobservable past with unique and non-repeating phenomenon. Science is incapable of proving it wrong.
But this is completely backwards. Creation is actually the account that’s easier to falsify. Macro-evolution on the other hand, is unfalsifiable.
For example, to falsify creationism, we just need to wait 8 million years and see if any more fish turn into mammals. If so, we have falsified current creationistic theories.
Macro-evolution itself, though, is not so easily falsified. Let’s say we wait 8 million years (I’m going to spend it on my front porch wearing a very heavy cardigan and smoking a pipe), and no more fish have grown legs and lungs. Humans haven’t grown scales and started laying eggs, apes have not developed civilization and culture, and pigs (still) can’t fly.
“Well,” says the macro-evolutionist, “evolution is evidently slower than we thought.” Eight million years later, given the same data (again), they can still say the same thing (again). There is not empirical piece of data that would prove the theory wrong.
Now I ask, which theory is more like the Witch Theory of Gravity?