RSS 2.0 Feed

» Welcome Guest Log In :: Register

    
  Topic: Advantages of theft over toil, The paper discussed< Next Oldest | Next Newest >  
Wesley R. Elsberry



Posts: 4966
Joined: May 2002

(Permalink) Posted: Sep. 16 2002,07:46   

William Dembski comments upon the paper John Wilkins and I wrote last year:

Quote
(1) Then why not withhold judgment in the Contact example and simply attribute a long sequence of prime numbers from outer space to unknown causes? The problem is that Wilkins and Elsberry's revised filter scotches all design inferences and not just the ones they don't like in biology. For the ID critic, the answer is not to revise the filter but to try to substitute a different picture of scientific rationality (e.g., Sober's likelihood approach). But that is deeply problematic itself.

(2) With regard to false positives, to say that the design filter does not commit false positives if there is specified complexity remains true. And to say that an attribution of specified complexity may be mistaken is also true -- and not inconsistent with the latter claim. There's a difference between specified complexity as it subsists in nature and our knowledge of it. You might want to reread my post about what sort of property is specified complexity.  


Thread on ISCID Brainstorms board

Dembski is incorrect in his assertion in (1).  Our revised filter does not eliminate all design inferences.  We went to some trouble to distinguish two classes of design inferences, ordinary and rarefied.  Ordinary design inferences are still just as valid as they ever were under our revised filter.  But the revised filter makes it clear that the epistemic warrant for rarefied design inferences is an illusion based upon invalid analogy to ordinary design inferences.

I find the point of (2) to be exactly what I've forwarded as a critique in the past, notably in my presentation at the "Interpreting Evolution" conference in 2001.  I have not been shy in saying before that Dembski's explanatory filter/design inference (EF/DI) is only reliable in the sense Dembski gives when one has complete knowledge, i.e., the true causal history is already known.  In that case, one has no need for Dembski's EF/DI -- it's entirely superfluous.  It is only in the case of limited knowledge that false positives become an issue, but these cases are also the only ones where Dembski's EF/DI could possibly have any utility.

It's nice to have Dembski confirm that I was right in making that critique.

--------------
"You can't teach an old dogma new tricks." - Dorothy Parker

    
  0 replies since Sep. 16 2002,07:46 < Next Oldest | Next Newest >  

    


Track this topic Email this topic Print this topic

[ Read the Board Rules ] | [Useful Links] | [Evolving Designs]