Replies

Well, my good humored General, consider the following. If we continue with the Hamlet analogy, we would have to believe that every "move" from the ur-Hamlet text to the authoritative text is "better" in some sense. So, we have to come up first with a sequence of textual changes each of which results in an intelligible text, and each of which is somehow "better" ("fitter"?) than the last. There is a well-known word game in which the challenge is to start with one word and by single letter changes move toward another word by stages, each stage being a meaningful word. It's hard--try turning "good" into "evil" using one-letter changes. What's hard is the constraint that each intermediate word be a real word and not nonsense. So, imagine the ur-Hamlet--and someone randomly changes a letter. How many of the possible altered texts is intelligible? Perhaps a few. How many would be considered "better" than the original? And how could you construct a path from the ur-Hamlet to Shakespeare's Hamlet? Daunting odds--a literary text's merits depend on the gestalt of the text--there may be sensible textual emendations on a small scale, but it is highly improbable that literary merit can be achieved by singular random changes--even if you have millions of years to fool around. Yet this is what the evolutionists would have us believe. I don't believe in a 6,000 year Earth, and am inclined to believe that variation is life forms is due to changed DNA. But as to why the DNA changed--here I part company with you. "Natural selection" has not been proved in my view, and frankly I doubt it could be proved.

I question your intermediary assumption that every single permutation of the ur-Hamlet need make strict literary sense. I think we can relax this assumption somewhat. How about if we restrict the intervening editions of Hamlet to something that the audience considers sensible? So, the following permutations are all considered equivalent:

To be or not to brie.

To be or not to bree.

To be or not to bhryea.

The audience will make sense of the word according to their training and socialization. As you are fond of pointing out the origin of the term "gook," I shall use that example to make my point.

The question is, then, how much does this relaxation of assumption buy us? Hmmm.

There is a well-known word game in which the challenge is to start with one word and by single letter changes move toward another word by stages, each stage being a meaningful word. It's hard--try turning "good" into "evil" using one-letter changes. What's hard is the constraint that each intermediate word be a real word and not nonsense. So, imagine the ur-Hamlet--and someone randomly changes a letter. How many of the possible altered texts is intelligible? Perhaps a few. How many would be considered "better" than the original? And how could you construct a path from the ur-Hamlet to Shakespeare's Hamlet?

Wait, wait. Both of you are over-restrictive.

Here's the problem - you're changing the terms of the analogy away from what it is to something else entirely. The point of the Hamlet example is that "fitness" is defined as closeness to the actual text. Now, by that standard, "To fly or not to fly" is less fit than "To by or not to by", because it's farther from the actual text, even though it arguably "makes more sense." But "making sense" isn't part of the original analogy, and the only effect of inserting such a restiction is to artificially pump up the odds against the final text being produced.

You're inserting a restriction that doesn't exist in the original, and for which I can see no good reason to insert at all - this idea that it has to "make sense" to an imaginary audience. Further, using single-letter substitutions, there's no guarantee that you can get from a given random string to the final product at all if each intermediate step has to "make sense". Finally, if we start with a random string to see whether we can get to the final product, you've set it up in such a way that we're dead in the water right from the start - what are the odds that some random string will "make sense"? Does the semi-random string (BX@# *&uyb 37(*FDb9s (WEJKhbec x3&^@b76&@vxb3 "make sense"?

Of course not. But by the original rules, it doesn't have to "make sense" to anyone - it just has to be a string that randomly mutates in order to move closer to the actual text, passing the mutations that are more fit from one generation to the next. Really, by what standard do we judge whether a particular genome "makes sense"? It doesn't have to be pretty, or elegant, or whatever - it just has to survive and reproduce. That's all - the Shakespeare example Dawkins proposes is already far more constrained than nature is. Dawkins has to get from "A" to "B" to produce a definite, fixed final product. Nature has no such constraints - it just has to go from "A" to "somewhere".

You've overloaded the analogy in such a way as to destroy its usefulness, and the final result is that you're not really beating up Dawkins's example at all, so much as your own construct.