Re: [asa] on science and meta-science

From: Iain Strachan <>
Date: Tue Nov 10 2009 - 08:19:11 EST

On Mon, Nov 9, 2009 at 10:40 PM, Keith Miller <>
wrote (inter alia)
 Secondly, human
> causal agents are NOT the same as divine agents.  This is a very important
> distinction.  Humans are causal agents that we can directly observe and
> study.  Other organisms are also similar agents (we can study the purposive
> behavior of animals and identify their past actions).   We cannot study God
> through scientific methods.

I think this is a very important point to make. When we infer design
of, say a watch, or a piece of architecture, we already know
independently that such designers exist - our inference is whether the
object of interest was indeed designed by one of these designers that
are known about. However, it seems to me that in ID, one is inferring
the existence of the Designer as well as inferring Design, because the
"evidence" of design is taken to be evidence of the existence of the
designer. But we don't need that final step with a watchmaker, or an
architect - we had independent evidence before. Keith mentions the
"argument from ignorance", which often folks take exception to. But
it goes like this: "I don't believe nature can do this unaided, so it
must be designed". But in the case of a piece of architecture, one is
not in a position of ignorance. I've seen something like this before,
and I met the architect who designed it. So I think it's a fair bet
that he or another architect designed this one.

Furthermore, God is unconstrained and can
> accomplish any logically possible end.  As I have argued on other threads in
> this forum, to be meaningful as a causal agent in science,  the capabilites
> of an agent must be constrained.  Otherwise an appeal to such an agent is
> identical to an appeal to ignorance.

This and the point above succinctly summarise my own problems with
Intelligent Design. If one just says "it must have been Designed"
then there is no constraint on the capabilities of the Designer. To
use an example from maths that I have used before. Suppose I have N
observations X at N different times T. Then mathematically I can
always construct a polynomial function that exactly predicts the value
of X for each of the N times T. All I need to do is fit a degree N-1
polynomial function (a0+a1.t + a2.t^2 + .... a(N-1)t^(N-1)). But to
do this says absolutely nothing interesting about my data because it
can be done for ANY set of N points. However, if I say that the set
of N data points is accurately modelled by a quadratic (ie degree 2)
polynomial, then I have said something of value, because not all
datasets can be modelled in this way.

Although, as Cameron has pointed out "design" and "miracle" are not
the same thing, they nonetheless have one important feature in common.
 They can both be invoked to explain absolutely anything. Hence they
explain nothing; just like my unspecified order polynomial explains


To unsubscribe, send a message to with
"unsubscribe asa" (no quotes) as the body of the message.
Received on Tue Nov 10 08:19:42 2009

This archive was generated by hypermail 2.1.8 : Tue Nov 10 2009 - 08:19:42 EST