Re: The scientific vacuity of Intelligent Design

From: Don Nield <d.nield@auckland.ac.nz>
Date: Fri Nov 25 2005 - 22:45:47 EST

Pim van Meurs wrote:

> Dawsonzhu@aol.com wrote:
>
>> Pim Van Meurs wrote
>>
>>> I am working on a larger posting addressing this excellent paper but
>>> let
>>> me point out that Intelligent Design is nothing more than the "null
>>> hypothesis" and can thus not even compete with 'we don't know'. In
>>> fact,
>>> we have seen in the past how a 'we don't know' position were used to
>>> propose evidence of a deity, only to be replaced later when our
>>> ignorance decreased.
>>
>>
>>
>>
>> I'm getting a little lost here. I would see most of the problems with
>> application of the "design inference" (in matters of experimental
>> science) to be that of establishing proper probabilities.
>> Ignoring the common (and useless) rhetoric from the ID group,
>> I'll go back to "The Design Inference". Although the last chapter
>> gets a little into the direction Dembski was intending to go, within
>> the body
>> of the book, the basic model was something like this: (1) __find__ a
>> correct
>> probability function (2) set a lower bound on the odds that are
>> reasonable to expect. The examples therefore centered on
>> card games (etc. ) where the odds can be predicted without dispute.
>> Even "honest" gambling parlors can make money off of these odds
>> and I suspect with a high degree of reliability.
>>
>> The major problems seem to appear when this is applied to empirical
>> problems where not all the information is know with certainty. I don't
>> quite see what you mean by vacuity. People use probability theory all
>> the time, and the issue should be which model is to be used and how
>> accurately it reflects what we observe in the world. Quite
>> understandably,
>> probability models immediately raise suspicions (as they should), but
>> that
>> does not make them vacuous.
>>
>> by Grace alone we proceed,
>> Wayne
>
>
> By vacuity I mean that ID does not add anything to our scientific
> understanding. For instance in the design inference, design is that
> which remains when chance and regularity have been eliminated. This
> however does not add any more to our understanding of a particular
> system than if we were to replace the 'design inference' with a 'we
> don't know'.
>
> Without any further constraints, the Explanatory Filter does not
> present much of anything scientifically relevant beyond having
> eliminated chance and regularity hypotheses, but this does not depend
> on the ID hypothesis itself. In other words, elimination of hypotheses
> cannot give scientific content to ID beyond what regular science
> already offers. In other words, we have to look at the conclusion
> drawin by the EF, namely that what remains should be considered 'design'.
> Finding the correct probability function is but one problem of false
> positives, finding relevant hypotheses of chance/regularity is another
> one. In other words, despite Dembski's claim that the EF is free from
> false positives, we do not really know, and have no way to establish,
> how common such false positives are.
>
> So now we have a 'design inference'. How can we compare this
> hypothesis with the already existing hypotheses or even the hypothesis
> of 'we don't know'? Let's assume that the probability for known
> processes is smaller than 150 bits, does this mean necessarily that
> the design hypothesis (whatever that may be) has a probability larger
> than it's competitors? We don't know because the probability of the
> design inference is never calculated, because it is inferred from the
> failure of 'competing hypotheses'. So why does the design inference
> deserve such a position?
> Why should we not accept 'we don't know' until we have independent
> evidence or a real hypothesis to be tested?
>
> Of course, the answer is that such a position would make the Design
> Inference utterly useless for testing supernatural events.
>
> In the limited ID relevant research, it already is self-evident that
> the design inference at most can lead to tests of existing scientific
> hypotheses and that ID never presents an independent hypothesis of its
> own, other than by arguing that the failure of other hypotheses,
> combined with a vague concept of specification can only be explained
> by intelligent design.
> Does such a position help us understand why? According to Behe we know
> that the designer wanted to design the object, and we know that the
> designer had the ability to design the object (Kitzmiller testimony).
> Even the former assertion is not clear as we know that intent may not
> be needed for something to be 'designed'.
> So ID does not give us much of anything that helps explaining the
> 'designed' object. Nothing about, how, why, by whom (motive, means and
> opportunity), no independent evidence (eye-witnesses), no alibis...
>
What Wayne has said about the design inference is reasonable as fars as
he goes. I agree with him that the problem is essentially that of
assigning the appropriate probabilities. But it precisely here where
Dembski meets a wall. In the case of card games the probailities can
indeed be assigned accurately, but in the case of biology Dembski can
never assign the probabilities, even within orders of magnitude, because
he does not have the knowledge -- and Behe will never be able to give
him the knowledge -- to assess all the possible alternatives. All
Dembski has is a Designer of the the gaps argument, and Pim is
justified in calling Dembski's argument vacuous. Another way of
expressing this is to say that the design inference is set up in such a
way that false positives are fatal to the ID argument, and Dembski will
never have the knowledge to rule out false positives with certainty.
Don

--
Received on Fri Nov 25 22:48:35 2005

This archive was generated by hypermail 2.1.8 : Fri Nov 25 2005 - 22:48:35 EST