On the theme of improbability...
Many people have difficulty with the concept of evolution, misunderstanding it to say that it's a combination of time and blind chance. One of the Hoyles (Fred?) said it was like imagining that a tornado blowing through a junkyard could accidentally create an aircraft.
It isn't blind chance, however, that powers evolution. It's selection. Variations between one animal and the next lead to unsuccessful variants being disposed of and successful ones goint on into the future.
Richard Dawkins in one of his books showed how the phrase 'Methinks 'tis very much like a weasel" could be 'evolved' in a very short number of iterations from an initial guess by retaining the correct letters from each guess until the whole phrase was arrived at. The same phrase, if you tried to arrive it simply by guessing and guessing until you matched the complete phrase would take eons.
So this nice guy has written a program so you can try it out yourself. Put any phrase you like in at the top, click the button, and it'll get there in a few sessions. (No, that doesn't apply to the complete text of the Gettysburg address - don't be silly!)
The problem with this (equally fatuous and equally wrong) example of how Evolution works is that it posits the a priori existence of a correct or optimal solution that somehow has simply to be reached.
On the airplane idea, it's a little like suggesting that an airplane with one wing and no engines (the rest having still to be evolved) can fly, just not quite as well as one that's fully formed.
Surely, in this case, the "correct" solution is the one that works? Or at least, works better than the other solutions?
If we know that something bestows an evolutionary advantage, then we can attempt to reverse-engineer the process, can't we? Or am I missing something?
But the ability to fly "a little bit" might help leap from tree to tree - and so on.
Except that in the weasel case, your test for what works is "what's closest to the original text". In nature, there is no original text. In the weasel case, when you've recovered the original text 100%, that's the end. In nature (leaving aside the suspicion that human ingenuity has put an end to evolution) evolution is a continuing struggle uphill - but facing backwards - so you can't see if you've reached any kind of summit or conclusion, or even if there is one.
I'm not saying that the weasel example is invalid - but it does ignore or misrepresent a significant concept in the way evolution operates by introducing a (drumroll please!) finish to the process.
The human genome has around 3.2 billion base pairs.
I wonder how many generations that applet would take to evolve a sentence with 3.2 billion letters.
O dear, El. Well below your usual standard.
To state that this doesn't model evolution very well is to state, like Sybil Fawlty's specialist subject on Mastermind - the bleeding obvious. I - and indeed Mr D originally - were quite clear that this is not a demonstration of how evolution works, but a demonstration of the difference between selection and randomness. As such, one must have something against which to match the selection.
There are currently 1 users browsing this thread. (0 members and 1 guests)
Bookmarks