Total Drek

Or, the thoughts of several frustrated intellectuals on Sociology, Gaming, Science, Politics, Science Fiction, Religion, and whatever the hell else strikes their fancy. There is absolutely no reason why you should read this blog. None. Seriously. Go hit your back button. It's up in the upper left-hand corner of your browser... it says "Back." Don't say we didn't warn you.

Monday, January 26, 2009


In mathematical modeling these is this little thing known as "identifiability." While it sounds, and is, highly technical, the basic idea is actually quite simple. In order for some model to produce a single unique answer in response to a set of data, we must impose certain constraints on it. These constraints, usually in the form of one or more assumptions, give the model a foundation or starting place from which it can do its magic. A useful analogy is to a game of sudoku- in order to figure out the complete table of values, you need those starting numbers. Without them, there are a very large number of possible configurations that would yield a solution or, in other words, it is those starting numbers that make a single unique solution identifiable.

This issue of identifiability has been on my mind of late because I've been thinking about Intelligent Design. For those who have been living in a dank cave without means of communication, intelligent design or ID is a supposed scientific theory that is essentially nothing more than Paley's famous watchmaker analogy embedded in a meaningless soup of pseudo-mathematical mumbo-jumbo. In accordance with my interest in the intelligent design nonsense I keep an eye on Uncommon Descent, the blog of Wild Bill Dembski and recently it featured a post which I find rather interesting.

The title was "Life on Mars, ID, and a prediction." Since a lot of you probably don't think about this as much as I do, I should mention a few things. First, this post on Uncommon Descent is discussing the recent evidence for the current, or past, existence of bacteria on the planet Mars. Second, intelligent design has often been observed to be non-scientific and, particularly, of being unable to yield a viable prediction. Of particular importance, it has often been claimed that ID cannot produce useful predictions without saying something about the nature of the designer. This last point the ID folks disagree with strongly, claiming that they are only "detecting design" and say nothing about its origins. So, the mention of a "prediction" is important for the ID folk because successfully generating predictions would be a major step towards legitimacy. Or, in shorthand, if scientific theories produce predictions, and ID produces a prediction, ID is science! Dembski wins!!!1!

Are we all clear on this?

Good. So, what is this prediction? Well, let's take a look:

So what does this mean for ID? Well, it means that those ID supporters who put stock in the notion of panspermia and directed panspermia are looking good. ID supporters like myself, UD author Doctor (MD) David Cook, and NASA physicist Rob Sheldon (see papers 45 and 46), are some of those. And of course I’d be remiss if I didn’t mention the discoverers of the DNA double helix Francis Crick and Leslie Orgel who authored articles and a book about directed panspermia.

I will now make a prediction from an ID perspective. Any living organisms found on Mars will be based on DNA and ribosomes essentially identical to what all life on earth utilizes. This is because life, even the simplest forms, is too complex to have originated in our solar system very early in its history. Wherever it came from, and however it got here, it was the same basic structural form that landed in all places - Earth, Mars, and wherever else in our solar system it may have found suitable conditions.

Right. So the prediction embedded in there is: if life is discovered on Mars then it will be based on DNA and related structures that are clearly related to what we have on Earth. This is, indeed, a prediction. However, there are two serious problems here- one empirical and one deeply theoretical.

The empirical problem is simply that this ID "prediction" is not unique. That is to say that if we were to rely on a model in which life emerged naturally on either Mars or Earth it is entirely reasonable to think that cross-contamination may have occurred. Likewise, Earth and Mars have reasonably similar sizes (Mars is smaller), reasonably similar compositions, and are in the same solar system. If life appeared both places independently, it did so under very similar conditions and thus may have similar structures. I suspect that there would be differences in the DNA, but the emergence of DNA on both worlds would not be totally shocking. So, in short, this ID prediction does not allow us to distinguish it from other theories. This is important because it means that even if there were support for this prediction, it would not support ID. By analogy, if police officers break into a room and find a woman holding a gun with a man bleeding from a stomach wound, they can be reasonably sure that she shot him. That information, however, is consistent with an accident, self-defense, and attempted murder. Without more information, they cannot distinguish between those options. Or, put differently, the theory is not identifiable based on the observation.

The theoretical issue involves the author's placing of constraints on the model. Specifically, he remarks: means that those ID supporters who put stock in the notion of panspermia and directed panspermia are looking good.

Panspermia, of course, is the notion that life arrived on Earth from extraterrestrial sources- for example bacterial spores drifting through space. Directed panspermia, likewise, is the notion that some intelligent agent deliberately spread such spores to generate life elsewhere. Now, the author doesn't indicate that this is a constraint but it functions as one. Specifically, it acts as a constraint because now we know something about the designer, and thus can generate predictions. So, for example, if non-directed panspermia is responsible for the emergence of life on Earth, then it is reasonable to think that both the Earth and Mars would have been exposed to the same spores. Likewise with directed panspermia- even if the directing agent visited us, it seems likely that they would have used life based on the same principles for the seeding of both worlds. In either case, by knowing something about the "designer" we know something about the mechanics of propagation and can therefore judge likelihood.

But now, let's think about what ID actually clams: that it doesn't say anything about the designer. The designer could be an alien race, spreading itself through directed panspermia or it could be an almighty god. But for just a moment, assume that it is an almighty god. What predictions does that generate? Well... none. The problem is that an omnipotent, omniscient, immaterial being can, presumably, do whatever it wants however it wants. If it chose to put life on Earth and Mars using the same type of stuff (i.e. DNA), it could. If it wanted to use totally different stuff, it could. If it wanted to create a really, really old looking universe in an advanced state, it could. If it wanted to create a young universe and actually let it age naturally, that would be okay too. The point is, this type of designer is consistent with any type of observation and, as such, no observation can be said to uniquely or definitively provide support for it. By extension, any time ID is used without a specific non-deity based designer being posited, the approach is always unidentifiable and therefore useless. ID without a designer is like Sudoku without starting numbers: pointless and entirely unsatisfying.

So does this mean that ID as specified in the post I refer to is identifiable? Well, no, it's not. There's actually nothing in the above prediction that is unique to ID or even requires an intelligent designer. Undirected panspermia doesn't involve any intelligent agency and, even if some intelligent civilization tried to spread life around the universe, they could easily have just loaded naturally-occurring bacteria on to rockets and fired it off. While intelligent agency, in that case, would have been involved in getting life to Earth, the life was not itself intelligently designed. The example of an ID prediction is, in fact, little more than a half-assed panspermia prediction.

What this example does do, however, is show that ID is entirely incorrect about not requiring a designer. In an ironic twist, when an ID proponent tried to derive a prediction, the very first thing he did was impose a constraint to try and make the theoretical model identifiable and, as it happens, that constraint was to specify the nature of the designer. In trying to prove that ID is scientific, the ID folks simply demonstrated that without specifying a designer, their perspective cannot generate useful predictions.*

Better luck next time, I guess.

* As it happens, it needs more than just a specified designer to produce useful predictions, but that's not the point right now.

Labels: , , ,


Blogger Arieh said...

The Theory of Evolution is science. It does not answer all questions, nor is it complete, but that is science - we are still learning and studying, and we always will be.

Intelligent Design is not science. It is philosophy, or maybe theology, applied to and drawn from science. Using science as a tool to improving our understanding is exactly it's purpose. Using science to back philosophy or theology does not turn philosophy or theology into science.

I hate people who confuse this. Religious nut-jobs who want to teach philosophy in a science class, and one-trick-ponies who think that thought should end with science and we shouldn't teach children anything else in any class ever.

... Actually, I think I just hate people n general. YOU ALL SUCK!

Tuesday, January 27, 2009 6:20:00 AM  
Blogger Drek said...

Crap, and I thought I was angry!

Seriously, though, I pretty much completely agree with you. Well, except for the hating people in general. I'm asocial more than anti-social.

Tuesday, January 27, 2009 7:10:00 AM  

Post a Comment

<< Home

Site Meter