Science on Trial (and Error)

So, P.Z. Myers is having a go at Bill Maher over various comments he’s made about science and health, particularly around vaccinations. The key annoyance Myers has is this:

But that’s not the part that had me fuming. It’s the bit around 4 minutes in, in which he pretentiously announces to us that not all science is alike, and climatology is a good science, so he accepts global warming, and he also explains that there is also consensus among climate scientists (he also argues that the earth is just a rock, so it’s simple enough to understand — but then, as he demonstrated so well on this night, Bill Maher is a goddamn idiot). And then he tells us that medical science is nothing like that, “because they’ve had to retract a million things”. “People get cancer, and doctors just don’t know why,” he says, condescendingly. His father had ulcers, and they treated it wrong when he was a kid.

So, what is Myers’ defense of science:

Science is a trial and error process. It is not an infallible track that leads invariably to correct answers, instantly and every time. When he says that climate science is completely right, that’s because he has only the most superficial knowledge — he knows a little bit about the conclusions they’ve reached now, but nothing about how they came to those answers. I guarantee you, there was a long slow gradual effort to understand climate, with false starts and dead ends and pointless detours all along the way, because that’s how science works.

When medical scientists retract something, it’s because they’re doing normal science. Of course there are errors along the way! The whole point of science is that you generate hypotheses, you make tentative conclusions, and you test them, and sometimes you’ll confirm your hypothesis (which means you’ve learned something), and sometimes you’ll falsify it (in which case you’ve learned something else). Do you know why Maher’s father got the wrong treatment? Because the cause of ulcers, the bacterium Helicobacter pyloris, was not discovered to be the causal agent until 19-****ing-84. Maher is complaining about an important medical discovery, one that won the Nobel Prize, because scientists didn’t discover it soon enough for him.

This … is not a good defense of science. It’s not a good defense of science when people point out that it’s gotten things wrong in the past — which should rationally weaken your confidence that the answer it is giving you now is correct — to say that of course it gets things wrong! That’s just part of the scientific process! Sure, science testing things, getting hypotheses wrong, and then correcting them is indeed an important part of what makes the scientific process good and something that produces knowledge, but the charge here really that we want and need to be able to tell when what science is telling us is right. And this is hard for science to do, at least credibly, because it has indeed gotten, well, pretty much every theory wrong at some point. At what point can we say that the answer that science is giving us is the right one, and the one that we should unreservedly adopt?

Scott Adams recently talked about dieting and nutrition:

What’s is science’s biggest fail of all time?

I nominate everything about diet and fitness.

Maybe science has the diet and fitness stuff mostly right by now. I hope so. But I thought the same thing twenty years ago and I was wrong.

He then goes on to list a vast number of things that people thought were accepted science about diet and nutrition that we all now don’t think is accepted, and in fact that people now think are settled in the precise opposite direction. When it comes to dieting and nutrition, beyond “Don’t eat too much of, well, anything” and “Eating clean fruits and vegetables is a good thing”, we don’t really know what sort of diet we should be eating. For now, we have issues with processed food, but given the history that might change in the future.

Adams turns this into a long discussion about why people don’t trust science, based on its history of having to adjust based on new information. The key is that we need to be able to tell when the science is settled so that we can trust and act on it, and in general science can’t tell us that the science is settled because the very nature of its process is that it pretty much never is. We can get a strong confidence in what it’s saying, but any new result might change that and, unlike other views, actually necessitate adopting a completely different theory to cover all the new cases, because science strives to generalize as far as possible. So even though we can say that Newtonian Physics works locally and you need Relativistic Physics beyond that, what that really means is that Relativistic Physics is the right theory and that you don’t need to consider the relativistic aspects at local levels, which is why we thought for so long that Newtonian Physics was right.

So if science sets out solutions as the settled and true answers and they then have to backtrack later and say that they were wrong, then science loses credibility. But if science does talk about answers as being provisional and subject to change then it doesn’t have the credibility to get off the ground in the first place; people wants answers, not answers that might be true because of a bunch of evidence that they can’t understand and evaluate and that might change later. Science needs to talk only about the science that is pretty much settled, and clearly distinguish between the parts of their theories that are settled and true and the ones that are more speculative. But often that distinction is not made, either by people who support science or by the media reporting on it. Any time that science says that something is true and not provisionally true and then has to retract that, it really looks like science just can’t figure out what’s really true and what isn’t … and defenses like “Well, science is critically trial and error” only make it look like that’s a key part of the scientific method.

Science produces knowledge, and is definitely the best method we have for figuring out truths about the natural world. What science is missing — and what philosophy of science needs to provide — is a set and solid epistemology that can set out when something is a scientifically confirmed theory that we can be confident in claiming to know, when something is “just a theory”, and when something is just a speculative hypothesis, and then needs to treat all of them in the appropriate way, from scientists themselves to those who philosophically want to use science to justify all truth claims to the media that reports on science. Without that, people will point to its failures as an indication that it is generally unreliable … and it definitely isn’t.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: