First off, I don’t think the two are mutually exclusive. What scientists need to remember is that the nature of their work is inductive.
For example, all of the observed oranges are orange, so we conclude that they are all orange. It’s not a terrible argument. But what if we find one that’s green? The assumption is, of course, that it’s not an orange.
Eventually, maybe someone decides to take it to the lab and test it. And they realize that it is, in fact, an orange. Now our set of oranges includes all orange oranges and this green one. Thus, the science of the color of oranges becomes that they are orange unless they are green.
I recently was debating online about the possibility of virgin birth. Of course, we have a model that explains how babies are made. Call it the birds and the bees, or whatever. But it’s impossible to actually observe every single conception. The problem is that a non-zero amount of women [0.5-1.0%] claim to having been virgin moms. Well this obviously doesn’t match the established model, so they must be liars. Right? Maybe. Chances are great that a large percentage of these women are lying. There is also the possibility that they got pregnant by sperm left on a toilet seat, or something else.
Let’s consider for a second that one in ten billion pregnancies are virgin births. Would we ever know? I don’t think so. Because the nature of science is to make this sort of inductive argument. They are great for describing most things. But if you experience something outside of our model, you must be lying.
I think we need to be careful not to use science to make false dichotomies. For instance, this is how all observed babies have been conceived. Thus, any babies claimed to be conceived other ways are impossible. It’s not terrible science, unless you have hundreds of women claiming to experience the ‘impossible.’ That’s when you check it out. That’s what science is for, to inspect truth claims. So why not test the genetic makeup of some of these children? This would make the scientific argument even stronger if they are all genetically normal.
But instead of using science, we give our opinions [based on what we know about science] to make the conclusion that all of these women are liars. The problem is that the truth does not care what your opinion is. And if that green fruit is an orange, you’d never know until you tested it.
So by believing that all of this science is true, and that anything that contradicts it is false, you may be throwing out important data. Which is not far from what science accuses religion of doing: using blind faith to ignore evidence.