Gildner Paleo Home Page
Life is still
|
The dancing letters at the top of this page demonstrate the power of evolution by natural selection. (It's not an animation, but a Java applet. You will see it only if you have a Java-enabled browser, and have Java activated in your browser's preferences.)
In evolution by natural selection, a new feature arises by random mutation. Mutation is inevitable: everyone, including you, have about 100 mutations. Most mutations have no effect on the owner. Some mutations are fatal, and some are beneficial. Beneficial mutations give the owner a better chance of surviving, and if the owner has a better chance of surviving, they have a better chance of passing the beneficial mutation on to their offspring, who in turn pass it on to their offspring, and so on. In this way, the genetic composition of a species can be changed over time. When enough changes accumulate, a new species has arisen.
In this demonstration, the "organism" is a word with 8 letters. Initially, the letters are random. The program has been written such that the letters in the word "evolving" each provide the organism with a benefit. Natural selection predicts that a word (and its descendants) with one of the letters in the correct position would have an advantage over others with a different letter. The letter is fixed into position, and further mutations affect other letters. In a very short time, the random series of letters evolves into the word "evolving."
Without the process of natural selection, this string would not quickly evolve into "evolving." Instead, all combinations of 8 letters would be generated, a total of 26 to the 8th power (over 200 billion) possibilities. By Natural Selection, preserving those words that are more fit, the process converges after about 200 possibilities.
( E-mail for more information on this Java applet: Raymond F. Gildner, macops@spacestar.net )
It has been pointed out to me that Dawkins refers to this process, using almost the same analogy in his landmark book, The Selfish Gene. His estimate of the amount of time it takes to converge on the "answer" is lower. The time to convergence is the number of letters times the number of possible letters; in this case, 8 * 26 = 208 generations. I've run the program literally thousands of times, and consistently average that number. As for who stole whose idea, I'm sure he's never heard of me, and I hadn't yet read his book when I created this program. RFG