withallyourmind.net

Over the next few posts, I just want to give some cursory and unorganized thoughts about some issues involved in the debate about Intelligent Design. By way of disclaimer, my views about these issues are not fully settled in my own mind, and I’ve done only introductory reading on most of the topics. My approach is philosophical, not scientific, and my dad and I have come to some different conclusions on several of these matters.

First I want to look at the question of how we come to believe that things are designed, and whether design beliefs are scientifically legitimate. A helpful way to do this will be to discuss William Dembski’s explanatory filter (which I will do in this post) and then talk about some problems it may have and what they mean for Intelligent Design (which I will do in the next post).

To start with, it’s obvious that we all detect design around us all the time and that, except in some rare circumstances, we are usually justified in doing so. One way to see Dembski’s explanatory filter is as a more formal (and therefore presumably more acceptable for scientific purposes) way to come to these same conclusions. According to Dembski, if we are looking at any particular thing, there will be three possibilities about the origin of that thing; either it came about by law, by chance, or by design. The filter is a method for deciding among these that “faithfully represents our ordinary practice of sorting through things we alternately attribute to law, chance, or design.”

So how does it work? Well, the first step is to see if the event of the thing’s coming-to-be can be shown be necessary or highly probable based on prior circumstances and the laws of nature. If so, then we attribute the event to law and not to design.

The second step is to figure out just how improbable the event was. If the event is only relatively slightly probable, then it is attributed to mere chance and not design. For example, if someone flips a coin four times and gets all heads, then that may just be attributable to chance. However, if the person flips a coin 500 times and gets all heads, then chance is no longer an option.

Finally, we must check whether the complexity inherent to the object or the event of it’s coming-to-be, now that we know it is sufficiently improbable, is specified. Defining what counts as specified gets a little tricky, but it has to do with whether the pattern exhibited can be meaningfully separated from the particular instance it is found in. A straightforward example is that of throwing 11 scrabble pieces on the ground. Suppose the tiles end up on the ground like this: “AJ EH OFSCA LK.” We would obviously attribute that to chance. Now suppose that instead the tiles ended up like this: “THIS IS A WORD.” Although this outcome is equally as unlikely (well, given some perhaps not plausible assumptions about what letters are available, etc.) as the previous one, the pattern that these letters ended up taking is specified in a way that they weren’t the first time. Therefore, we can conclude that this outcome was somehow designed, and is not explicable by chance.

That, in a nutshell, is Dembski’s way of justifiably inferring design by ultimately looking for specified complexity. Read his article at http://www.arn.org/docs/dembski/wd_explfilter.htm for a fuller and perhaps more understandable treatment. Next post, I’ll say what I think about the filter.