Weaselution

Intelligent Design object lessons - show evolution's issues as complexity grows

Richard Dawkins' Weasel program demonstrates evolutionary processes—random variation combined with non-random cumulative selection. Asserting that given enough time, small incremental genetic code changes produce complex systems. It is a simplification based upon the ‘infinite monkey theorem’; given infinite time, a monkey and a typewriter could eventually create all of Shakespeare’s work. The Weasel program extracts a small phrase from Hamlet, "METHINKS IT IS LIKE A WEASEL" as the evolutionary target phrase. It begins by choosing a random sequence of 28 letters, duplicates it repeatedly, but with a random chance of 'mutation'. The program examines the mutant phrases, the 'progeny' of the original phrase, and chooses the one which most resembles the target, METHINKS IT IS LIKE A WEASEL.

Weaselution mechanizes this and expands upon it - allowing you to make your own evolutionary assumptions. You can experiment with the number of progeny, the mutation rates, the target phrase and most significantly, the complexity required for selection.

In Dawkins' weasel program, every character which matches the target is considered a benefit and immediately selected for the next generation. In reality, things are more complex - a partially evolved feature would be of no benefit; so that partial feature has no better chance of being selected than any other and will 'devolve out'. A more correct analogy is that complete words are required to be useful and selected for the next generation. Weaselution demonstrates complexity implications by allowing you to select one of 4 complexity modes:

mode 1: none - every target phrase matching letter is assumed to be selectable by nature. But Why would an unfinished word have any selectability? Some complexity is required for the feature to be selectable, just like you need at least a word to have meaning.

mode 4: whole word - assumes some complexity is needed. For something to be selected (better fitness) it has to perform a function. This is simulated by not allowing a word to be selected until the whole word is in place. MELDINLS scored as 0 since the whole word is not complete. METHINKS scored as 8 since the whole word is complete. This increases the odds significantly that a word will devolve before it can be selected.

mode 2: Complexity with buildup/fixed position - A common rebuttal to irreducible complexity is that it assumes that components are useless on their own. In the evolutionary model, the sub components of large assemblies can do other jobs, hence they are selectable. Weaselution models that concept in this mode. Complex functions (longer words in our sim) are made up of smaller less complex functions (shorter words). For example: Methinks sub words: Me, thin, in, thinks, I, ink, think. It uses the rule that smaller words must exist in Hamlet to be considered a valid subword. (the Weaselution contains all of Hamlet and verifies subwords exist in that work). This mode includes a key cheat – sub words are forced into the correct position in the larger word. Example: in Methinks: Me------, not ---Me---

mode 3: Weaselution Demo 4: Complexity with buildup/unspecified position. Same as mode 2 except the subword’s position is not pre-specified. The subword could be selected but be in the wrong position to create the larger word. Subwords have to assemble themselves in the right order by chance.

Use Weaselution’s analogy to demonstrate evolution’s short-comings. It can be used as an effective apologetic tool for intelligent design. Especially when you consider the HAMLET simplification as compared to life. For example: the average word has 5 letters, this produces a 1 in 10,000,000 chance of getting all letters in place at once. The smallest protein, insulin, has of 51 amino acids, this produces a 1 in 10^66 chance of assembling a useful protein by chance.