Thursday, July 30, 2009

So very wrong

I was over at the wired science blog (which is very good by the way) when I ran across this in the comments on one of the posts.

There is no big surprise here. What most people consider “Science” is often glorified Voodoo Artistry. Even the “hard sciences” like Physics and Math have plenty of unsolved mysteries. The soft sciences (Biology) have plenty too. Stuff like Oceanography and Meteorology (silly science) are almost totally librarian, i.e. they are really detailed observations without much of what Scientists consider “science”, that is, they have almost no predictive power.

It should surprise very few hard thinking people that much of what mass media considers “science” is often lame. How is it that so many “smart people” have bought into the Global Climate Crisis when our “weather scientists” can hardly predict (accurately) the weather a few weeks out. And, “they” expect us to believe that 20 years out we will be subject to catastrophic climate change.

Here’s the deal : lots of people who study “science” want to be considered “scientists” so they can feel good about their work. Maybe we should consider a stronger definition of what we can call science. If your body of work cannot accurately predict 99% of it’s claims, it cannot be considered science. If your work has no definable experimental framework, it is not science. If your work opens more questions than it answers, it is not science.


... and no reading the original post doesn't really help. The original post is here and was on some anomalously high tides they've been having on the east coast. It would be difficult to fit more misconceptions about science into such a short space. I addressed some of the problems in a response in the comments. But it seems a shame to keep it buried in some comments so here it is.

Ok, I really don’t have the time or inclination to address what you have written in any detail, and I rather suspect it wouldn’t matter just because your mind seems to be made up no matter what the facts are. But for the benefit of anyone reading who might actually think you know what you’re talking about let’s cover some things. (and yes I did this very quickly and I’m sure there are typos… sorry)

”How is it that so many “smart people” have bought into the Global Climate Crisis when our “weather scientists” can hardly predict (accurately) the weather a few weeks out. And, “they” expect us to believe that 20 years out we will be subject to catastrophic climate change.”

Weather versus climate, they’re not the same go learn the difference

“lots of people who study “science” want to be considered “scientists” so they can feel good about their work.”

While, unlike you, I cannot speak for “lots of people who study “science”” I can speak for myself and a few of my colleagues. I can honestly say that how we feel about our work has nothing to do with what you or anyone else calls us.

“If your body of work cannot accurately predict 99% of it’s claims, it cannot be considered science. “

Did you read that sentence after you wrote it? How do you predict a claim? I’m Paleontologist. If I claim that an asteroid killed the dinosaurs how then, do I predict it? I might predict we’ll find a crater but if we don’t am I wrong or have we just not found it yet?

But let’s back off of the semantics and look at the broader point. You seem to be claiming that science should be predictive. To some degree it should be, but 99%? The hallmark of science it not that it’s predictive, it’s that scientific ideas are tested. This means that sometimes we get stuff wrong. When that happens we get rid of what doesn’t work and, hopefully, replace it with what does, then that idea is tested etc etc. What this means in the cumulative sense is that we get better at stuff. By your definition if, right off the bat, if we can’t make predictions with 99% accuracy then, apparently, we’re not doing science. If I desperately cared whether of not you were calling me a scientist then the solution would be to simply gather data until I could predict things with 99% accuracy. Two problems: First predictions are very useful at well below your 99% benchmark. Hurricane track predictions are nowhere near 99% accurate (depending on how you define accuracy) but they’re still very useful. If we wait to be 99% accurate we’ll be waiting a very long time. And in the case of hurricane track predictions people will be dying while we wait for that 99%. Second, we actually learn a lot by predicting stuff and being wrong, that’s how we test our ideas. If we never predicted something and were wrong how would we know whether or not our ideas were correct? Learning what doesn’t work is sometimes as valuable as learning what does.

“If your work has no definable experimental framework, it is not science.”

Why? I’m a paleontologist. I do some experimentation but most of my data are collected in the field. Why are data collected through observation in the field less valid than data collected through experimentation?

“If your work opens more questions than it answers, it is not science.”

Once again: why? So if your answer to one question leads to more, of it your work shows is that we really don’t understand something as well as we thought we did then it’s not science? Once again if we’re unwilling to take on what we think we know then how will we ever know if what we think we know is wrong?

Update: He responded in the comments. Go read if you want it's.... interesting?

No comments: