Richard Dawkin's big blunder

[Esquilax]

If evolution is blind, then shouldn't cumulative selection be demonstrable without a target? This is a fair question.

How would one go about proving a negative like that? That's the problem I'm having; to ask if this is demonstrable sans target would mean that you're presuming you've seen evolution happening naturally with a target, something that needs to be demonstrated in itself.

Do you agree that for any selection criterion, there will exist some set of potential outcomes which evolution will home in on?

That's a difficult question though, is my point: convergent evolution demonstrates that there are certain outcomes that are obviously better suited to a given environment, so in that respect the answer is yes. On the other hand, there is no guarantee of this, and it's entirely possible that, given the random nature of mutations, we would see evolution down a path that's merely good enough, rather than the seemingly "optimum" path convergent evolution shows.

To say these things will be homed in on is to apply a level of direction here that simply isn't present; things slip through the cracks all the time, and evolution isn't a binary beneficial/nonexistent process. All we can really say is that there are certain traits that have a proven track record of working, and that if they do show up they should be favored. But evolution is weird, and the interactions within its framework aren't as easy to predict as you're wanting to make them.