Re: [asa] Doug Groothuis v. William Dembski

From: Rich Blinne <>
Date: Thu Jan 01 2009 - 11:52:45 EST

On Jan 1, 2009, at 8:08 AM, Iain Strachan wrote:

> On Thu, Jan 1, 2009 at 2:53 PM, David Opderbeck
> <> wrote:
>> In fairness, most ID advocates do often cabin their arguments a
>> probabilistic, but they seem to act like they have acheived absolute
>> certainty.
>> The other important difference between a design inference in nature
>> and the
>> work of a forensic scientist in a criminal case is that criminal
>> trials are
>> all about human conduct. Here we get into the theological
>> questions: to
>> what extent is there an analogy between God's attributes and the
>> creation
>> sufficient to draw an inference of divine design from an artifact
>> of nature
>> that looks designed to a human.
> I think in the above paragraph you put your finger on the whole
> problem with the intelligent design inference. I don't think the
> question of "absolute certainty" is the issue here (after all, even
> Dawkins doesn't claim to be absolutely certain that God doesn't exist
> - he is content with "almost certainly").

I think we need to be careful that we don't accuse ID of something it
isn't trying to prove. The "probabilities" are not for the existence
of God but rather whether the evolutionary mechanism works or not. The
problems is not the use of probabilities per se but that ID is
completely incompetent when it enters the area of probabilistic
analysis and information theory. Note the cluelessness that resulted
when Randy tried to explain what "random" meant in the January 2008
issue of Christianity Today. (See
  Both information theory and population genetics are robust fields.
Yet, ID attempts to hijack both. When experts in these areas cry foul,
ID should take notice. There is an important reason why peer review is
called *peer* review.

For example, there was an impossibly high probability assigned that
the blood found at the scene wasn't OJ's. If ID had something like
that then they would have something significant. The problem is they
pull their "probabilities" completely out of the air. In addition to
establishing guilt looking at alleles (see
  can also determine paternity. These are measured from people who are
related. By extension we can compare the genomes of other living
things and see if they are related. The problems that result when you
actually measure the probabilities is that it shows common descent and
that goes in the complete opposite direction that ID wants.
Furthermore, ID should take a clue here of how real forensic science
overcomes a common problem that they have, the lack of prior
probabilities. When a cladistic analysis is done it often uses
Bayesian analysis. (see
Bayesian_analysis) Bayesian inference of phylogeny is based upon a
quantity called the posterior probability distribution of trees, which
is the probability of a tree conditioned on the observations. It uses
Markov chain Monte Carlo (see
  to approximate the posterior probabilities of trees. Note to the
uninitiated: when you see the phrase "Monte Carlo" think randomized
trials. The name was chosen because of the use of dice at casinos
found in Monte Carlo.

Real forensic science does use probabilities but they are measured
ones and are not pulled out of the air or done through ill-formed
concepts such as specified complexity. As we see above the issue
raised by David O., the lack of prior probabilities, is not an
insurmountable one. Thus, the detection of design is theoretically
possible despite the current incompetence of the ID movement. ID wants
to claim the cache of forensic science because of the claim of a
common use of "probabilities". The problem is they do not seem to have
the chops necessary to really understand the underlying concepts
whether the concept is "random" or "information theory". This is the
reason why they are generally disrespected by secular (and believing)
scientists and not because they are theists. If they want to be
forensic scientists they need to be real forensic scientists and not
just play ones on CSI: Waco.

Rich Blinne
Member ASA

To unsubscribe, send a message to with
"unsubscribe asa" (no quotes) as the body of the message.
Received on Thu Jan 1 11:53:20 2009

This archive was generated by hypermail 2.1.8 : Thu Jan 01 2009 - 11:53:20 EST