subscribe to our mailing list:
|
SECTIONS
|
|
|
|
Title |
Author |
Date |
Criticisms of Dembski's latest opus |
Wilson , David |
Aug 19, 2004 |
I was flattered, but rather taken aback, to see myself described by Mark Perakh as an "expert" in fields pertinent to the paper William Dembski recenty posted at the URL http://www.iscid.org/ubb/ultimatebb.php?ubb=get_topic;f=6;t=000533 . While I do have a sufficiently strong background in all the fields necessary to follow the mathematics of the paper, it would be an exaggeration to call me an expert in any of them.
It also not clear to me whether the errors I found in the paper are of any real importance. Although Dembski does refer to the property which my counterexample disproves as one of "the requisite properties we have come to expect from an information measure", he makes no use of it at all. Also, judging from Shalizi's criticism, the failure of Dembski's "variational information" to satisfy this property doesn't appear to have prevented it from having found wide application as the Rènyi information divergence of order 2. Finally, as far as I can see, there is no intuitively compelling reason why it should be required to have this property.
In short, while the errors are ones which I should be embarrassed to have made myself, they aren't any more embarrassing than some which I actually have made myself on occasions. Thus, athough it's certainly worth pointing them out, I shouldn't be inclined to make much of a fuss about them.
|
|
|