| Problems with Dawkins' Weasel: |  |
|
Richard Dawkins used his considerable intelligence to fashion an algorithm demonstrating Naturalistic Information Generation.
His program starts with a random sequence of 28 letters or spaces, then generates copies with 'errors' representing mutations. The program checks 'offspring' sentences, and selects the one closest to the target sentence, allegedly representing natural selection. This supports his claim (in The Blind Watchmaker) that evolution by random changes, combined with selection, is virtually inevitable.
Dawkins' Weasel is representative of amino acids forming a functional protein, where each letter (or space) combines to make an English Sentence just like individual amino acids combine to create a functional protein. However, this simulation contains three glaring problems:
1. The DNA of most life-forms code for 20 different amino acids (some code for 21 or 22). In fact, the DNA code for a specific amino acid sometimes differs from one species to another. (Another problem for common ancestry). Also, there are more than 50 amino acids that occur naturally, so Dawkins' Weasel correctly allows his simulation to start with letters not found in the target sentence.
Amino acids found in nature can be either 'left-handed' (L) or 'right-handed' (D); the description for the orientation of the molecule. This distribution is roughly 50-50, and no known naturalistic process can separate them.
However, only 'left-handed' (L) amino acids are used to create the proteins found in life. (Amazingly, only 'right-handed' sugars are used to create DNA nucleotides.) This means that there are twice as many amino acids to choose from when generating our "sentence".
Weasel Casino corrects this problem by including both upper and lower case letters from which to randomly choose (representing both left and right handed proteins).
2. The main thrust of Dawkins' Weasel is to demonstrate an example of an 'Evolutionary Algorithm'. This algorithm would be a simple formula (which must occur naturalistically), but when given the right input, could generate great complexity and even useful information.
However, the amount of information or intelligence generated by an algorithm can never be greater than the intelligence or information that went into creating the algorithm. Since, as has been stated above, Richard Dawkins is very intelligent, his program must be disqualified as an example of an 'Evolutionary Algorithm'. If only Dawkins' Weasel had been the result of random coding.
Weasel Casino corrects this by eliminating the algorithm and reflecting the real world where a protein must be functional and complete the first time, or it is discarded. Biology insists: "Get it right the first time." Remember, "Dice have no memory."
Of course biological and colonial algorithms do exist, but their ingenuity and pragmatism strongly indicate a need for their own designer.
3. The use of a 'target sentence' forces a direction on a system, which is supposedly demonstrating how no direction is needed.
The intelligence needed to determine whether we are 'going in the right direction' is completely anathema to the precepts of methodological naturalism, since there is supposedly no one to indicate what 'right' is.
-
Jon Saboe
Founder: Evolution Is Dead