Real Science is Hard

When I started this series, I said I wasn't going to go through the book systematically, but I guess I am. I think to understand Stokes' approach, I have to provide at least some insights into each chapter. The next chapter is pretty long and gets into considerable detail about how science works...and it's a bit hard to understand; thus the title of this post. I'll try to cover all of it in one post.

According to Stokes, science has brought humans tremendous advancements and...it works! But just because it has been successful in many areas of human endeavor, this doesn't mean that it is right about everything. In this chapter, Stokes demonstrates that even where it shines - in the world of physical reality - it gets things wrong: "And if science likely gets things wrong about earthly things, why would we believe it when it comes to heavenly things?" While the scientific method is often presented as clear-cut, objective, and fool-proof, we'll see that it is more complex than that and it is as human as every other human endeavor; that is, it is prone to human quirkiness.

Inference is the key to the scientific method and is at the heart of theory-making. When something is inferred it is not directly observed. If a friend of mine comes into my window-less basement and he's soaking wet, I'll infer that he was caught in a rain storm even though I didn't see the rain or him running through it. Of course, it's also possible that he ran through a sprinkler or went swimming fully-clothed, but the most plausible explanation based on my life's experience and the limited data I know about the outside world (rain was in the forecast) is that he was caught in the rain. So, when we talk about inference and theories we are attempting to explain - why do things appear this way. A theory is an explanation. And as you can see in my simple example - inference to the best explanation - is not always straight-forward; there are typically other explanations. In general, then, we can say that science is in the business of inferring theories. Stokes sets up the rest of the chapter with this:

But no matter how well a theory predicts or represents observable phenomena - no matter how well it is supported by this agreement with observation - it is still a theory, simply because it is inferred and not observed. Not only is evolution 'merely' a theory, but so too are general relativity and quantum mechanics, the most well-attested scientific theories humans have devised. The real question is this: how good are our reasons for believing them; how good are our inferences?

I guess this will take more than one post. See you next time.