Down at Overwhelming evidence, Denyse O’Leary — yes, again… — wrote a somewhat ‘interesting’ post (in the clinical sens of ‘interesting’, of course) about design inference. I would have liked to react faster to this post, but I was stuck to my desk by those stupid evolving platyhelmynthes of mine. Screw you, evolving thingies.

According to her claims, 9/11 is an exemple of design inference. When you saw the first airplane hit the tower, you probably thought that it was an accident. Just when the second airplane hit the second tower, you ceased considering this option, and realized it was terrorism. That, people, is design inference explained by Denyse O’leary.

For a short moment, I said to myself “Maybe she’s right, and this is a design inference”. It actually is. But it is not — in any way — transposable to natural evolution. 

Why? It’s all about statistics – or maybe probabilities. According to Denyse O’Leary, and many other ‘design inference’ enthusiats, species can not emerge by chance. Juste like two airplanes just can’t hit two adjacent towers by chance, and it was later proved that they didn’t : the crashes were designed. 

The flaw in the logic here is that airplane crashes in 9/11 where not independent. It was only two manifestations of a single decision. Even if two planes hit two towers, it is nothing but a single event. In evolution, events are more likely to be independent. Since two populations are unable to exchange genes — i.e. to maintain gene flow — they stop to be dependent (unless being involved in an ecological interaction, but this is a particular case). 

I think there is a deeper flaw in Denyses’ logic. We now know for sure that there was a designer behind 9/11. We were able to connect the pattern (plane hitting tower) with the process (someone taking the decision to attack New-York). In this case, we know both the conclusions, and the premices. 

But it is dangerous to apply the same logic to evolution : of course, if we consider that our premice is ‘there is a designer’, every single attempt to infer design will lead us to the conclusion that ‘there is a designer’. Any serious evolutionnist will start by investigating the premices (even if in certain cases, it is usefull to investigate patterns in order to determine processes, which is close to indudtivism).

That is the main problem with “intelligent” design : they try to conclude to the existence of a designer, knowikng that the existence of this designer is their only hypothesis. Where the heck is intelligence in this?

Also, see this tremendously good post of Riley Firth, at The War on Bullshit.