>> One must grant that two copies of the same information produces no
>> increase.
>
>No, one doesn't need to grant that. Within Shannon's theory, two copies of
>the same sequence yields 1 bit more of information than when it is alone.
>As Yockey said, information is a mathematical object. One can't use
>semantics to deal with it. One must calculate it.
>
>glenn
>
>Foundation, Fall and Flood
>Adam, Apes and Anthropology
>http://www.flash.net/~mortongr/dmd.htm
>
>Lots of information on creation/evolution
The "information" that Dembski is dealing with isn't really Shannon information, although it is related. You are right, though, that a useful theory must recognize that multiple copies of the information are significant, because the extra copies can be modified without losing the original information. Dembski fails to take account of that aspect, as far as I can see. I have only read the central chapters that contain his core argument.
A few other comments on the core argument:
1. Someone who is being hyped as "the new Isaac Newton" should know what a conservation law is and isn't. Dembski's "conservation" law (of complex specified information [CSI]) as described is a "stay the same or decrease" law, not a conservation law. But if you look at the development of his argument, he admits that "small" amounts of information can be accumulated by chance. So his law is really a "stay the same, decrease or increase a little" law. As an aside, this "law" seems to translate in Phillip Johnson's usage to an assertion that no information at all can arise from other than intelligent sources, which is simply false. It is trivial to take a yeast ura- mutant that lacks the information to grow without uracil and let it revert to a sequence that can grow without uracil. The difference may be only one base (2 bits), but it now has the CSI to grow, whereas before, it didn't.
2. Dembski's assertion that CSI cannot be created by chance depends on his further assertion that it is impossible to accumulate small amounts of information in increments to reach the level of CSI. But this appears to be a baseless assertion. It simply begs the essential question.
3. Dembski and his cohorts need to face the fact that the relevant information is not just a mathematical abstraction, it is protein and nucleic acid sequence. They need to break down and learn some physical biochemistry. The question of whether new CSI can be reached by accumulation of small increments depends on whether selectable biochemical function can be achieved at an adequate frequency in random sequence proteins. Dembski treats CSI as something that can be assessed qualitatively (the correct info is either there or it isn't.) But protein function is a quantitative matter. Even very weak enzymatic activity can support slow growth, and that makes improvements in the enyzme possible by further selection . Behe has made some arguments that activity would not occur in random proteins at an adequate rate to be significant (http://www.leaderu.com/orgs/fte/darwinism/chapter6.html), but he has been answered, I think rather effectively, by Ken Dill. To follow in Glenn's tradition I include the abstract below.
TI - Polymer principles and protein folding.
[Review] [103 refs]
SO - Protein Science 1999 Jun;8(6):1166-80
AB - This paper surveys the emerging role of statistical mechanics and polymer theory in protein folding. In the polymer perspective, the folding code is more a solvation code than a code of local phi/psi propensities. The polymer perspective resolves two classic puzzles: (1) the Blind Watchmaker's Paradox that biological proteins could not have originated from random sequences, and (2) Levinthal's Paradox that the folded state of a protein cannot be found by random search. Both paradoxes are traditionally framed in terms of random unguided searches through vast spaces, and vastness is equated with impossibility. But both processes are partly guided. The searches are more akin to balls rolling down funnels than balls rolling aimlessly on flat surfaces. In both cases, the vastness of the search is largely irrelevant to the search time and success. These ideas are captured by energy and fitness landscapes. Energy landscapes give a language for bridging between microscopics and macroscopics, for relating foldi
ng kinetics to equilibrium fluctuations, and for developing new and faster computational search strategies. [References: 103]
Dill points out that the relevant target for a random leap into protein sequence space is not a single point or even a contiguous region. It is all the regions that could fold to some structure(s) that would have the necessary activity to any degree, and there are probably multiple unrelated structures (regions of sequence space) that would do the job in most cases.
However one evaluates the arguments of Dill and Behe, it seems to me that this is the relevant turf for the argument on whether CSI can accumulate. The question is approachable experimentally, although it is technically difficult. Of course, these arguments assume that cells exist as vehicles for the selection of functional proteins. The origin of cells is a whole 'nother kettle of polymers.
This archive was generated by hypermail 2b29 : Fri Apr 21 2000 - 21:44:48 EDT