>Well, considering that there were no feathers in the fossil record in rocks
>earlier than this, and there are feathers in the next geologic era, gives
>us a clue as to which direction the transformation was occurring.
We just don't know that there was such a transformation. The possibility
exists that the scales of L. are independently evolved from some precursor
of feathers.
>OK, where are the complex spines in the fossil record? What animal has these?
The fossil record is pathetic fragment of history. There must be theory, not
just to interpolate, but to puzzle out yet unknown processes.
>>I take seriously the principle of reduction and specialization of skeletal
>>segments. I'd be okay with Ichthyostega as the ancestor of all terrestrial
>>quadrupeds, if this animal had enough bones in the digits to evolve into
>>primates, for example, through a process of reduction. But it doesn't.
>
>au contraire, it does. THis is from my web page and you can verify what I
>say in the references there:
I was talking about the number of bones in individual digits, not the number
of digits. But it doesn't matter much. If it wasn't ichthyostega, it was
something
similar, perhaps a smaller arboreal version.
>I don't think many would argue that we can never know the animals that are
>directly upon the line of descent. But we can get close.
What is "close?" Either you have a fossil or you don't.
>Woah. I didn't say I believe in 'gradual' evolution. I believe that
>punc-eq is a much better model. It fits with nonlinear systems which is
>what life is, and it fits the data of paleontology better.
Isn't PE still gradualism? It's mere common sense that there should
be periods of stasis. There's no real insight into processes here.
>Not always forever, there are whales that are born with feet, there are
>occasional snakes born with feet, there are occasional horses born with
>three toes as their ancestors and Caesar was said to have ridden a 5-toed
>horse.
Sure, there are atavisms. But we are talking about great trends. The tides
are what they are, despite eddies here and there. Eddies don't grow into
new tides.
>Occasionally things can come back. Dollo's law really is a probability law
>that says it is extremely unlikely, but not impossible.
What things? Dollo's Law began with observation.
This is all much clearer if one accepts the principle of reduction and
specialization
among serial homologs, that is, among skeletal parts.
If we accept this principle we must posit ancestors with many symmetrical
segments, whose evolution is a matter of reduction in number of segments
and distortion of the remaining segments. Obviously, random variation acting
upon an array of symmetrical elements will produce loss and distortion of
the elements; random variation generating new symmetrical segments is
practically speaking impossible.
--Cliff Lundberg ~ San Francisco ~ cliff@noe.com