How Science Works, Part 1: What We "Know"

Hello. Where have I been? Hiding, that's where!

Not really. I've just been busy, doing research, pulling together manuscripts, doing illustrations, and being homeless and poor. Now I am more or less stationary in Salt Lake City for a while, I feel for the first time in a year like I have something resembling a routine, so I want to start writing again.

I've recently become interested in creationism again. One particular subject that I'd like to address is that of what science is actually trying to do, and what scientists mean when they talk about "proof." It is often argued that nothing can be "proven" in science, and that any claim to absolute certainty makes science a form of religion. The argument that science in general or evolutionary science in particular is a type of religion is a common one, but since it wasn't elaborated on, I won't get into it here; AronRa has a perfectly good dissection of it anyway. I want instead to focus on the statement that science doesn't "prove" things. I've talked before about how science is fundamentally based on the recognition of human fallibility, and want to elaborate a bit about what this means for knowledge, and what we can actually "know."

Science certainly does not “prove” things beyond any doubt, and there is always the provision in science that ideas, even well-accepted ideas, can change. The most extreme version of this philosophy, advocated by Karl Popper, is that a single piece of falsifying evidence can destroy an entire theory, no matter how many tests that theory has previously survived. This is an oft-repeated bit of bullshit which deserves an entire post, so I won't get into what I call "the Myth of Single Bullet Falsification" here.

The more general claim that any idea can be shown to be wrong is certainly true. However, the same thing is true of real life. That doesn’t stop us from forming opinions, and being reasonably certain about them. I’m pretty sure my mother isn’t a robot. She looks and acts human, and everything I know about modern technology and artificial intelligence indicates that building a robot as human-like as my mother is impossible. That’s the evidence I have at my disposal, so that is what I base my opinion on. And I am pretty confident about it, so confident that I would stake my life on it if I really had too.

Of course, I STILL could be wrong. There might be some secret lab somewhere that can make super realistic robots that no one knows about. And if I saw a reputable news story about such a lab, or caught my mother changing her batteries, I would revise my opinion. However, I have not been presented with that evidence. Right now, it exists only in my imagination. The real world evidence I can point to is what I listed in the paragraph above. So for the time being, I’m sticking with that. This is what scientists usually mean when they use the word “proof”; not absolute certainty, just certain enough that they don’t have any real doubts about it (for now).

There are different levels of certainty of course. I’m pretty sure (though not as sure as I am about my mother not being a robot) that my car will run tomorrow, that I will not be diagnosed with cancer tomorrow, or that a comet will not hit the Earth tomorrow. My current plans revolve around these things being true…but I could be wrong. Indeed, I'm much more likely to be wrong about these things than I am about the existence of lifelike androids. That doesn’t stop me from making plans. It doesn’t stop MOST people from making plans, most of the time, and when people are so incapacitated by possibilities that they cannot make rational decisions, we tend to throw around terms like "neurosis" and "mental illness." Part of being a sane, reasonable, and functional human being is whether or not you can distinguish between probabilities, possibilities, and pure imagination. The same thing is very much true of science, where there are also many levels of certainty. The basic difference between speculation, hypotheses, and theories has to do with how much evidence they have supporting them (speculation has the least, theories the most), and how certain we can therefore be about them.

Anyway, the point is that being really sure about something doesn’t mean that you can’t change your mind, and accepting that you can change your mind doesn’t mean that you can’t be really certain. We all juggle this conundrum in our day-to-day lives, and seem to do just fine. I’m really sure Mom isn’t a robot. I’m also really sure I could change my mind if the right evidence was presented. This may seem like a contradiction, but it doesn’t keep me from having an opinion or making specific plans about whether to take to a doctor or a mechanic if she is feeling poorly. The same thing is true of science. We can’t really “prove”, but we can be certain enough. I’m about as certain that evolution is a real thing, and that the Earth is more than 6,000 years old, as I am that my Mom isn’t a robot.