Richard Dawkin's big blunder

[Angrboda]

Argument from ignorance.

Its not an argument of ignorance. Its a probability argument. Let me explain.

If I have a bag containing three marbles each of which is either white or black what is the probability all the marble are white? To figure this out let W stand for white marble and B stand for black marble. There are four possible configurations

1)W-W-W
2)W-W-B
3)W-B-B
4)B-B-B

Principle of indifference applies here so each has a probability of .25. Now what happens if I randomly draw a marble and it is white? Now the probability that all the marbles are white increases to .33 because configuration 4 is no longer possible. What happens if I draw another marble and it is white? Well the probability that all the marbles in the bag are/were white increases to .5. Everytime a white marble is drawn while no black marbles have been drawn...it increases the likelihood all marbles are white.

Just for completeness, the odds of drawing white and black marbles depends on how many of each are originally in the bag. Assuming there are three white and three black marbles, the probability that all marbles will be white is not 0.25, primarily because there are more ways to draw a mixture of black and white marbles than there is to draw all white or all black marbles, so the drawing of a mixture of black and white marbles have greater than 0.50 probability and therefore neither all white nor all black can have 0.25 probability. Depending on how many black and white marbles are being drawn from, this will generally be the case. And the probability that you will draw all white or all black marbles does not increase with having drawn two of either; if there is a plentiful supply of both in the bag, the odds are always whatever the ratio of white to black is in the bag. Assuming equal amounts of both, the odds are 50 / 50 for drawing white or black. (The reverse of this, that drawing two whites increases the odd's of a black, is known as the Gambler's fallacy because it isn't true.) If there is not a plentiful supply, the possibility of drawing a black or white marble is dependent on how many of each are in the bag. If we again assume that there are three of each, after having drawn two white marbles, there is one white marble left, and three black ones, so the odds are 1 in 4, as there are 4 marbles in the bag, and only one of them is white.

I look at Darwinian evolution much the same way. I don't know if it white(the product of an intellect) or black(not the product of intellect). However each time I see an evolutionary system whose origins I know and it turns out that it is the product of an intellect, that in my mind increase the likelihood that all evolutionary systems are the product of intellects.

I don't know that this is so much an argument from ignorance as it is simply incorrect. For this to be true, you'd have to know that the evolutionary systems you've examined are more or less like the whole pool of evolutionary explanations; this is known as being a representative sample. Unfortunately for you, until you've drawn a sufficient number of systems from the pool of all systems that exist, this is something you don't in fact know to be true. As noted above with the marbles, whether you draw a black or white depends not on what you've previously drawn, but on the percentage of the pool that is composed of each. What changes as you draw more is the confidence that what you've already drawn is representative of the greater pool, and that only happens after drawing a great number of examples out of the total pool. Even then, this is what's known as an inductive argument; it doesn't lead inexorably to the right conclusion; it can always turn out to be wrong. Just as at one time, only white swans were known to exist, and it was therefore presumed that black swans do not exist, your so-called "guided evolutionary" systems could turn out to be the white swans, and naturalistic examples of evolution could betray that consistency and turn out to be black swans. As long as you're not examining natural evolutionary systems to determine whether or not they are guided or not, the amount of non-natural evolutionary systems you examine is basically irrelevant, as there's no reason to assume the pools of both, guided and naturalistic evolution, are composed the same.

This is a basic problem with inductive arguments in general. They frequently don't lead to useful probabilistic arguments.