Monday, July 16, 2007
Information Systems as Evidence of Design
I forgot to put in the reference list at the end, so here are my references:
- Abel and Trevors, Journal of BioSemiotics -- "More than a Metaphor: Genomes are Objective Sign Systems"
- Gitt, W. In the Beginning was Information (condensed form here)
- Shannon, C. The Bell System Technical Journal -- "A Mathematical Theory of Communication"
- Trevors and Abel, Cell Biology International -- "Chance and Necessity Do Not Explain the Origin of Life"
- Yockey, H.P. Computers and Chemistry -- "Origin of Life on Earth and Shannon's Theory of Communication"
First of all, the other levels aren't "made up". Gitt's information theory is a synthesis of other information theories already available, the only inventions by Gitt is the apobetic level and the pragmatic level (which actually probably already existing in philosophy anyway). Syntax and semantics are well-defined fields in language and computer science studies. To call these "made up" means basically trashing the work of Noam Chomsky and those after him. Syntax is taught to every computer science student, and semantics, as a discipline, is taught to upper-level students.
As to your question itself, I think it shows that you don't understand what statistical information means. Statistical information only deals with the size of the information pipe to move the information -- it has nothing to do with meaning. For example, if phrase "I have a dog" has a very well-defined semantic meaning. "asdklfjaklsdfjasdlkfaklgjewihgiasdnblnadlasdkfjaklsdjflkasdjdgladjsnlbjdas" does not. However, the second example has quite a bit more statistical information than the former. Adding statistical information, if not done _according to higher-level patterns_, removes higher levels of information, it does not add them. Now, if your mutational system understands, is aware of, and operates according to the higher levels of information, then you can do things which according to some metrics may be qualified as adding information. However, I am not aware of any case where one level of information can contribute to a higher level of information. This means that in order to have a system that generates information at a higher level, it must be built from a system of an even higher level. This indicates, at a minimum, that the origin of life must have been a very, very complex organism, rather than a simple organism.
I am currently in a dialogue with one of the Talk.Origins contributors on random mutations, and hopefully at the end we will publish our correspondence on some website or another. We cover this sort of thing extensively.
Sure they are. They are a contrived series of definitions that are not employed by anyone else in the actual field for the purposes of arguing against evolution.
Gitt's information theory is a synthesis of other information theories already available, the only inventions by Gitt is the apobetic level and the pragmatic level (which actually probably already existing in philosophy anyway). Syntax and semantics are well-defined fields in language and computer science studies
But not genetics? How about that... Silly me - I forgot that language analogfies and computer-based theories are universally applicable...
To call these "made up" means basically trashing the work of Noam Chomsky and those after him. Syntax is taught to every computer science student, and semantics, as a discipline, is taught to upper-level students.
They are made up in terms of how they deal with biological reality. None of what you mention is related in any way directly to biology. Its application is only via metaphorical language.
As to your question itself, I think it shows that you don't understand what statistical information means. Statistical information only deals with the size of the information pipe to move the information -- it has nothing to do with meaning.
No - that was my whole point.
Since the meaning of a DNA sequence is only derived from its sequence, then altering the sequence will alter the 'upper' levels of informaiton since those upper levels are dependant upon the statistical level. It has to be. Claiming that the upper levels are unaffected by such changes is charlatanism.
For example, if phrase "I have a dog" has a very well-defined semantic meaning. "asdklfjaklsdfjasdlkfaklgjewihgiasdnblnadlasdkfjaklsdjflkasdjdgladjsnlbjdas" does not.
Not to you, but suppose I were a spy and used such apparent gibberish to hide communications? And why the analogy? Let's see some real data from you people for once.
However, the second example has quite a bit more statistical information than the former. Adding statistical information, if not done _according to higher-level patterns_, removes higher levels of information, it does not add them. Now, if your mutational system understands, is aware of, and operates according to the higher levels of information, then you can do things which according to some metrics may be qualified as adding information. However, I am not aware of any case where one level of information can contribute to a higher level of information. This means that in order to have a system that generates information at a higher level, it must be built from a system of an even higher level.
Nice circular argument.
That is what Gitt's claims amount to.
Let's stop with silly contrived analogies, shall we?
An insertion is an event which adds statistical information to a genome, would you agree? According to you, unless the higher levels of 'information' somehow produced the lower level increases, then there would be no real increase in information (but that is not a contrived claim, no sir!).
Yet a simple statistical increase via insertion can produce alterations in the 'meaning' of the substrate in biology. This is documented. An insertion event that disrupts the promoter of a gene has produced an increased expression of a particular protein. This overexpression has conferred DDT resistence to a population of fruit flies. The 'meaning' of the gene has changed and it in fact now performs two functions - the protein is part of the electron transport chain, which still operates fine, and the excess protein uses DDT as a substrate and decomposes it. That is 'added information' in anyone's book. If it is not, then the whole application of communications system informasiton theory to genetics is a misapplication.
This indicates, at a minimum, that the origin of life must have been a very, very complex organism, rather than a simple organism.
Whatever...
Hmmm... so by your definition, since evolutionary biologists don't argue against evolution, anything which anyone uses to argue against evolution is "contrived". That's convenient.
"But not genetics? How about that... Silly me - I forgot that language analogfies and computer-based theories are universally applicable..."
They are universally applicable. The problem is that the assumption that DNA does not have its source in intelligence has made it difficult for geneticists to view the language aspects of DNA. It does not mean that they are not there.
"They are made up in terms of how they deal with biological reality. None of what you mention is related in any way directly to biology. Its application is only via metaphorical language."
Actually its related via logic and mathematics. Of course, in truth, everything related mathematically is related via metaphor (Math is actually just a very precise way of making metaphors - so if you don't like metaphors does that mean you don't like math?)
"Since the meaning of a DNA sequence is only derived from its sequence, then altering the sequence will alter the 'upper' levels of informaiton since those upper levels are dependant upon the statistical level."
It will alter them, but not in any specific direction. Adding statistical information may destroy the amount of information present in the upper levels. It may increase them. On the other hand, removing statistical information may add information. The quantity of statistical information has nothing to do with the amount of meaning in the final message.
"Not to you, but suppose I were a spy and used such apparent gibberish to hide communications?"
So what? It is true that you could write a code around it, but that's irrelevant. The point is that randomly generated information does not add information to a given code.
"Nice circular argument."
How is it circular?
"An insertion is an event which adds statistical information to a genome, would you agree?"
Yes.
"Yet a simple statistical increase via insertion can produce alterations in the 'meaning' of the substrate in biology."
I don't disagree.
The problem with your whole example is not that it's wrong, but that it's not random. The cell is using existing information to make modifications. That's perfectly fine. Insertion sequences and other mobile genetic elements are not random changes. They are part of a system that is built to produce variety. That is not against creation theory in any way, shape or form.
<< Home