Hacker News Reader

Comment by vijayboyapati | Hacker News Reader

The main problem with "climate science" is epistemological. It does not hold up to the foundational standards of the natural science which rely on falsification. With a system as complex as the earth's environment, falsification is essentially impossible. You cannot rewind the clock back 50 years, change a variable and play out what the temperature would be now. In a way, this makes climate science similar to economics. Yet climate "scientists" continue to act as if their work is as repeatable and falsifiable as a highschool lab experiment.

View on HN

If you take a very narrow view of falsification, then evolution is unfalsifiable too, and therefore not science (this is actually a common argument made by anti-evolution people).

That is untrue since evolution has indeed been studied in falsifiable experiments.

https://en.wikipedia.org/wiki/Experimental_evolution

On the other hand studying how exactly the current specieses evolved is more like history.

There are certainly specific experiments that have been carried out w.r.t. mechanisms (also true with climate science, fwiw). It's more the overall "big" conclusions that are a bit difficult to replicate in a laboratory, e.g. "the earth is warming as a result of human-caused greenhouse-gas emissions" and "humans descended from other animals through a process of evolution driven by natural selection".

Any examples of such experiments for climate science? Other than actually predicting climate change and then waiting 30 years I have trouble imagining falsifiable experiments. For evolution you can isolate a couple of populations with short generations in the lab and study them.

I agree with that argument.

It's history. If not for the whole religion dust-up we'd probably call it natural history.

Pretty much all science lies on a spectrum from, say Math on one end, thru physics and biology and eventually psychology (say) on the other. You take the view that some sciences are OK and others not, but it's more nuanced.

The greenhouse effect is solid physics and experimentally falsifiable and verified. We can measure fluctuations in solar output. But sure, the system as a whole is horribly complex and we can't make precise predictions.

Just as a matter of common sense we know that pumping crap into the atmosphere is probably a bad thing -- we shouldn't need a rapidly impending crisis to persuade us to try to avoid doing it, but getting people to do or stop doing anything is politics, not science.

Climate "scientists" make a big deal of pointing out that their work is simply a "best guess", and also trying to quantify the uncertainty. We know the greenhouse effect is real. We don't know how much damage (if any) it will do, but in order to make decisions about what to do, we need some kind of model.

It's not the lack of falsifiability that's bad - hard science can only answer some kinds of questions - it's the pretense of falsifiability that's a problem. Granted, you basically acknowledge this in your last sentence.

Which scientists are you referring to? There are thousands and thousands out there and I find your sweeping generalisation a little unlikely to say the least. I don't usually see hard short term predictions, it's more often presented as ranges of temperature averaged over significant periods of time.

I'm curious. Do you also consider astronomy, cosmology, paleontology, ecology, and geology to be "science" rather than science?

Natural history.

There is no magical standard for falsification. If we want to be 100% rigorous, then to have an experiment for any law of gravitation we would need to test every pair of bodies at every possible distance.

While "falsification" is impossible for complex models, the predictive power of a model can be shown to be inadequate. It takes years for it all to play out, but it's still science.

Economics is far worse since we have such a poor understanding of psychology.

I'm not sure which problem is the "main" one, but this is certainly in the top three.

No, the main problem with "climate science" is that it's impossible to take the political (and nearly religious to some on the left) aspects out of it.

Only then, can there be a serious discussion about "global warming".

"Nearly religious to some on the left" — um, I have no problem with describing some of the left's view of climate change as "nearly religious" if you accept that many on the right take a completely and utterly religious view.

You mean, literally religious. They do. On the contrary, calling the "global warming" belief a religion is metaphorical. However, I think it is a very good metaphor, as it immediately brings a certain enlightenment to the whole argument.

It is very unfortunate, in my opinion. As unfortunate as the politization of science.

Reply on news.ycombinator.com