Wednesday, December 31, 2008

Critical Thinking

It is common enough these days to be told that such and such is true, let us say for example a statement like "Mankind is destroying the planet by creating global warming." Now if you don't happen to believe that and present quite clear and logical arguments against the proposition that human activities are contributing significantly to the problem (given that it even is a problem) you are cast in the role of a "denier." It is rather like Chicken Little crying "The sky is falling" and if you say, "I don't think the sky is falling" you are accused of something like, "You are a fool if you think it's not" or "Your obstinacy is keeping us from taking effective action." Now before you take action of any kind you ought to establish that the thing is happening and if it is that the action proposed will be effective in producing the intended consequences.

But that's not my point. My point is that the discourse should never reach this point. The whole essential activity called thinking and more specifically critical thinking has been short circuited by a variety of rhetorical and senseless, frankly shameful fallacies that have kept any thought at all from happening.

Claims made without evidence are called assertions. An assertion "The sky is falling" is not an argument even if it is the assertion "Man is causing global warming." Calling someone a name is also not an argument it is a logical fallacy called ad hominem, and saying that someone is funded by the oil companies and therefore everything he says is wrong or suspect or whatever is another fallacy. A man paid by oil companies is as likely to be as honest as a man paid by politicians. There are long lists of fallacies which one really ought to have at least some kind of knowledge of.

But really all you have to know to think reasonably well is: 1) Assertions are not arguments, 2) Any assertion only becomes an argument when data, principles, observations, values and consequences are all brought to bear. Observations lead to data, and linked to data principles lead to expected consequences which can be confirmed. Values are also linked to principles which relate the values to the consequences in an evaluative way. Only then can we find any "oughts" that may be involved. "We ought to save the planet if it is endangered and if we can." If we discovered that the sun was going nova next week there isn't very much we could do except calmly prepare to die. But if we insist on living in New Orleans then it's sheer madness not to make the levies as strong as possible so that even Katrina sized storms won't breach them, or perhaps we ought not to build cities below sea level.

The problem is that much of our public discourse carries no data. We are being called upon to believe people because they make claims and among those claims is that they are scientists. The last time I looked, being a scientist myself, I noticed that all the scientists I know are fallible human beings like the rest of us. Virtually all the global warming brouhaha for example is based on computer models. I used to write computers models and still do now and then. I used to joke that GIGO (Garbage In Garbage Out) didn't mean that, it meant Garbage In, Gospel Out. That's because the aura of having come from a computer still carries some of the magic elixir of infallibility.

The case is very different. A computer model is never better than its validation and climate models are not only short on validation they actually have error bars associated with their various inputs that are larger than the variation in their outputs. In other words they are little different from speculation. You can't make a good computer model from components that are themselves poor models. It's the same way with critical thinking. A chain of thought is only as good as its weakest link. One bad component makes the whole chain of thought weak.

I remember reading Richard Feynman's very good little book The Meaning Of It All in which he comments that if you have a good argument that's all you need, but if your argument isn't very strong then you need to multiply reasons. That's the same principle that goes into making rope. A single strand may be quite weak, but if you multiply the strength with many strands it gets stronger. One very powerful strand would be enough (the strong argument) otherwise you need lots of weaker arguments that all point in the same direction. Note that I said arguments not conjectures, not assertions. Arguments have to have reasons that are backed up by data, real data, convincing data.

Nowadays it is rare to see any data at all. Instead we are told of consensus (an argument from imputed authority — that's a fallacy) or warned of dreadful consequences, or told that this is so dangerous that even if it might be wrong we have to act as if it isn't. This is all nonsense and fear mongering. What we need is less passion and more critical thinking. It's hard to do with a population raised on sound bites and spin.

No comments: