It’s been suggested that a “theory” is supported by evidence, and a “hypothesis” is just an idea that hasn’t been tested yet. These are sort of in the right direction, but miss the mark. And this has implications for discussions about catastrophic anthropogenic global warming, or CAGW.
A more strict definition: A hypothesis comes about from observing the world, developing a guess that explains the observations, making predictions that support the guess, then testing/observing further to see if those predictions hold up.
Those testable predictions must:
- distinguish your hypothesis from other competing ones, and
- be capable of being falsified (in other words, finding a result that shows your hypothesis to be wrong or incomplete). That’s okay, as you can go back through the cycle again until you get it right.
If a prediction explains any result, then it explains nothing, and you’re just spouting philosophies, notions, hopes and beliefs, however well backed-up you might be by models and logic.
Even science is not as strict as this ideal definition, when it comes to the terms theory and hypothesis. In the ideal, both would have some supporting evidence, but a theory would be a hypothesis that has been well-tested by successful prediction. Unfortunately, various science disciplines use “theory” for notions that haven’t quite risen to the level of a proper hypothesis yet, such as “string theory.” It’s not necessarily evil, just sloppy.
Thus, “theory” can be used to encompass things that are felt to be true by some group of scientists, but so far have not held up to rigid testing. And the broader and more complex the area covered, the tougher it is to develop a rigorous test.
This is where climate science largely is at the moment. The Earth’s extremely complex climate system, hugely fed by input from the Sun (more than 99.999%) and subtly modified by magnetic and other aspects of the surrounding space, is poorly understood. We don’t even understand the inputs, yet, let alone what happens after those inputs reach our biosphere. We’re just learning more of how the Sun’s changes affect cloud formation, and how the UV frequencies vary by tens of percent during the Sun’s usual 11-year cycle.
And, of course, the predictions (invariably of gloom and doom) are either unable to be falsified, or unable to be told from natural variation, or both. If they predict that Antarctica will be colder, and it isn’t, well, they predicted that too. Oh, it actually is colder? Well, they said that all along!
Anthropogenic global warming would be falsified, they said, if there was no warming in five years. No, ten years. Twelve! Fifteen! Sometime in the ambiguous future! But in the meantime, Global Warming is Real!
No more snow! Ah, extreme snow? Well, that’s “not inconsistent” with our predictions!
In short, such guesses are worth very little.
And positive indications, the actual increase in crops and so forth directly attributable to increases in CO2, get little airplay and little interest in the journals. The FACE (Free Air Carbon Enrichment) experiments have run for decades now, and have demonstrated great success from crops to trees, but they take these results and suggest that weeds will do better than edible crops, or that “popcorn will no longer pop and beer will be poisonous” as Nature reported in August of 2007.
The saddest part of all of this is that progress in some areas is being opposed and corrupted by the desire to come up with the “right” answer. Scientific American had an editorial a decade or so ago in which they said that if you offered a paper that suggested that global warming was not a crisis, they’d “laugh in your face.” We’ve seen, over the years, the selection bias that results from this distortion of the scientific method.
We have had models of physical systems for centuries, which are increasingly complex. In the process, we’ve gone from notions too simplistic to be useful (as improved observations showed) to models complex enough to no longer be understandable themselves, while still only a tiny sliver of the complexity of the real biosphere.
And we run those models, see if they produce the “right” answers, then tweak and run them again until we get what we want. All published models have gone through this; their “raw” output based on “logic” never sees the light of day until many (sometimes thousands of) iterations later, when the modelers like the result.
From this sort of filtered and distorted output, amplified by a supporting blogosphere, well-meaning and intelligent people have gotten the idea that there is an impending crisis, that the reality of massive food increases (due in part to CO2) are harbingers of food shortage, and that the pleasant (for humans) biosphere of recent decades portends doom for us all. Earth’s billions of years of feedback be damned!
Much of climate science, while describable loosely as “theories” in the colloquial usage that has even been adopted by scientists of various disciplines, is still far from being constrained by proper testable hypotheses. And such constraints are actively resisted, for a variety of motivations.
In so many cases, a given climate scientist might get ambiguous results, but he has been told all his career that others show clear proof of global warming — so he interprets his own work that way, not realizing that so many of the others are in the same predicament. It’s not dishonest, but it is poor science. This motivated me six years ago (long before ClimateGate) to rewrite a poem from Geoffrey Saxe that originated as “The Blind Men and the Elephant.”
It is interesting to me how much my poetic jibe matched the later-revealed reality, with scientists’ emails showing them unable to reproduce the work of others — and fearful to raise the issue. I didn’t mention, though, the fierce hatred of anyone else attempting to do that double-checking. Here’s an example, of both the corruption of certain scientists and of serious problems with the data — some of which may be accidental, some obviously not.
===|==============/ Keith DeHavelle