Re: [asa] philological notes on randomness (was: Re: What my tiny little brain was thinking...)

From: Rich Blinne <>
Date: Sat Nov 14 2009 - 22:47:00 EST

On Nov 14, 2009, at 6:47 PM, Schwarzwald wrote:

> I agree that "truly random", "truly unguided", etc is unscientific,
> and not what any scientific theory does or can say (not while also
> remaining 'scientific' at least). But I think Cameron has made a
> very strong case in the past that as far as Darwin was concerned,
> these things were *essential* to his "theory" - such that, if you
> strip out these claims (on the grounds that they are not
> scientific), you're left with a theory that, however useful, is not
> what Darwin was driving at. Not by a longshot

It's not essential to the theory. In population genetics the time
sequence of inherited genes are modeled as a Markov chain which is a
random process where all information about the future is contained in
the present state. As I stated before science defines random
differently than the OED. Here random means unpredictable in detail
though likely predictable in its statistical properties. A computer
program like PSI-BLAST finds the hidden Markov states and reconstructs
the common descent. (One of the key evidences for macroevolution is
BLAST reconstructions based on genetics give the same answer as BLAST
reconstructions based on fossils.) Another way randomness is used is
to test for positive selection. Normally what you see is random drift
in the genome. If you don't see random drift that indicates that gene
was selected. Here random means no correlation and not no purpose.
When ID tells scientists that evolution is random they laugh because
it is when you see non-randomness is when evolution via positive
selection is occurring.

I use Markov chains in my job. That's because I design hard disk
controllers and we have to extract the signal out of the noise. We use
what's known as a Viterbi algorithm which uses a MCMC (Markov Chain
Monte Carlo) algorithm to extract the real set of symbols from the
noisy ones. The question is why do I use Markov chains? Because it
goes all the way back to 1948 when Shannon created information theory.
Here he applied random Markov chains to a sequence of symbols. (A
clue that randomness is involved here is that information content is
related to entropy.) ID theory uses Shannon extensively. In the same
way as evolutionary theory ID is dependent on randomness because they
both use Markov chains. What this means is that ID theory just proved
that the genome is truly random and unguided! Of course this
conclusion is positively silly but is the result of conflating the
popular understanding of random with the scientific one. You can
rightly cry foul with the claim that ID was trying to show that the
information in the genome was truly random and purposeless, and I can
rightly cry foul when you do the same with evolutionary theory. If the
ID proponents would just get off their anti-evolutionary high horse
for a moment they can see that they could use science and evolutionary
theory against the New Atheists. Namely, the conclusions of truly
unguided evolution is neither scientific nor in keeping with current
evolutionary theory. Miller knew and knows it and is why he testified
the way he did in Dover. Miller and not Dawkins reflects what
mainstream science believes about evolution and is why the NAS and
AAAS have stated that evolution and religious faith can be compatible.
Miller is also right that the Levine misread Gould vis-a-vis
randomness and made a critical mistake.

Rich Blinne
Member ASA

To unsubscribe, send a message to with
"unsubscribe asa" (no quotes) as the body of the message.
Received on Sat Nov 14 22:47:45 2009

This archive was generated by hypermail 2.1.8 : Sat Nov 14 2009 - 22:47:47 EST