Half-truths and information theory
According to Darrell Huff, it’s easy to lie with statistics. But it’s just as easy to lie with any kind of science, for the same reason - there’s a lot of knowledge out there and most people aren’t familiar with the basics. So people can come up with intuitively plausible ideas like homeopathy using some quantum-theoretic means of healing, and many people will accept it.
Creationist lecturers make a living from this ambiguity: giving just enough knowledge to tempt the audience into making a leap to a fallacious conclusion. One of the examples the Intelligent Design creationists like is something called Information Theory, which they misrepresent and abuse readily.
Information Theory sounds like it might be about meaning and purpose. Of course, information in that sense is entirely in the eye of the beholder. (Imagine having a crossed line on the telephone while you’re talking to a friend. As far as you’re concerned, the other conversation going on is ‘noise’, while yours is ‘information’; the other people think the exact opposite.) There are no formalisms for deciding how useful something is. But the creationists don’t clarify what is meant by information; they just let you make assumptions because it suits their purpose.
In information theory, a message has information if you can’t predict what it says before you read it. So a dice with a single spot on each face provides no information - you knew it would show a 1 before you rolled. An ordinary dice is unpredictable, so when it stops you’ve learned some piece of information. This is how information theory regards information: as something which informs.
Whether it is interesting or useful is irrelevant. In fact, the most informative message possible is one which consists entirely of random numbers, since it has the most unpredictability.
As well as not defining what is meant by information, creationists then claim that information cannot be created. (This is why they bring it up: to show that complex DNA cannot evolve.) But from the correct definition given above, we can see it is very easy to create complex information. Any noisy signals - leaves rustling, background radiation - are fantastic sources of information.
They mean to say that structured, interesting information cannot be spontaneously created, so genomes require a designer. But in this regard, information theory is not helpful, so they lie about what it says and hope to create a convincing argument from the half-truths and fancy mathematical terms.
In further posts I’ll talk about the good stuff in information theory, and places where you’ll come across it. Tune in next time!