Well, the word de-evolve isn't really in my vocabulary so I have trouble following your arguments. I'll try and clarify what I am talking about, though.
It seems to me that teleology can be reconciled with a genealogical view of life in two ways (not exclusive):
1.) There is a purpose driven selector which chooses different variations each generation in accord with its final goal for life
2.) Variation each generation isn't random but is being directed towards an ultimate goal.
With regard to the first view, I was using an analogy with artificial selection and using Dawkins Biomorph program to illustrate my point. An applet can be found at:
Atheism v. Theism
Each generation, the user selects the variant with best approximates the final design desired, and that variant acts as a template for the next generation. This would be one way in which teleology can be realized.
So there would be two factors that would contribute to the "fitness" of a design, i.e., its chance of being around in the next generation: 1. How well it is adapted to the environment, and 2. How well it conforms to the plan of the "selector". If this is the case, it would seem that there would have to be some designs in nature which are sub-optimal for reasons which cannot be explained by developmental constraint or constraints the environment puts on the organism's design.
But also, it would seem unlikely that there would be large periods of little or no change. For even if a design is well adapted to the environment and any new variation would be selected out by stabilizing forces of natural selection, it wouldn't be perfect with regard to the purpose or plan of the selector and therefore change should still be occurring within the lineage.
Now with regard to directed mutation or variation, I think that could be demonstrated by direct observation. That is, we should notice variation having bias towards a particular design over another. I dont' think we see that.
This issue is very confusing and I haven't come close to exhausting all possibilites. I think this arises because the nature of teleology is so vague, i.e., what plan or purpose is there? Until that can be articulated it is very hard to say whether teleology can be tested or not.
Edited by humoshi, : Grammar