Re: Felsenstein v. Dembski

On Mar 30, 11:38 am, Seanpit <sean...@xxxxxxxxx> wrote:
On Mar 29, 8:32 am, hersheyh <hershe...@xxxxxxxxx> wrote:



On Mar 29, 11:16 am, Seanpit <sean...@xxxxxxxxx> wrote:

On Mar 26, 8:09 am, hersheyh <hershe...@xxxxxxxxx> wrote:

No.  What Felsenstein states is that *if* the assumptions of complete
randomness of "winner" sites in a completely random landscape *were*
true, then NFL would prohibit useful natural selection.  He then
points out that that assumption is not true and, as he concludes,
although the description of NFL is true if the underlying assumptions
about the landscape were true, it is irrelevant to the situation of
real evolutionary landscapes.

Not beyond very very low levels of functional complexity.  At higher
and higher levels of minimum size and specificity requirements, the
potentially beneficial targets become exponentially rarer and do in
fact take on a fairly uniform distribution throughout sequence space.
These are demonstrable facts.  These facts completely undermine
Felsenstein's main point and improve Dembski's basic idea.

You're notion, as well as Harshman's, that the landscape of sequence
space somehow stays highly clustered without any sort of uniform
distribution or exponential decrease in potential target sequences is
completely contrary to the known facts and is therefore based on
wishful thinking at best.  

The landscape of useful functions that *have* been discovered by
living organisms are, in fact, clustered into related families.

The problem is that these clusters are uniformly distributed
throughout sequence space and become more and more widely separated
within that space at higher and higher levels of functional
complexity.  No known novel systems of function which require greater
than 1000 fsaar are within striking distance via the mechanism of RM/
NS this side of trillions upon trillions of years of time.  You have
produced no examples - not one.

Yes, I know, you have produced some attempts at examples, but your
examples are either originally derived from the end product in
question, or they are the result of a loss, not a gain, of a novel
structural system - i.e., a loss of a pre-established system or
interaction.

 And
the pattern of relationship is, in fact, that which would be predicted
by a mechanism of descent with modification involving both neutral and
selective changes.  

You mean the NHP?  The mechanism of RM/NS can indeed produce NHPs and
can explain the similarities of such a pattern quite well.

And can do so not *only* for proteins that serve the *same* function,
but also shows linkage with similar proteins that perform *different*
functions. It is no accident that myoglobin and the two hemoglobins
show a pattern consistent with historical divergence. And actual
experimental evidence shows that the amount of difference between
aldosterone and cortisol receptors is a very few aa's. And that the
likely ancestor binds both steroids.

 However,
the mechanism of RM/NS has really problems when it comes to explaining
the functional differences.

What makes you think that?

In short, arguments for common descent are not the same thing as
arguments for the mechanism of RM/NS - -  not at all.

They are *exactly* the same. RM/NS leads to a pattern of descent with
modification. In fact, randomness alone leads to a pattern of descent
with change. In the real world you will get both changes that affect
function (and invoke positive or negative selection) and changes that
do not (which will lead to drift).

That landscape of *real* relationships is what
needs to be explained.  The fantasy landscape of inventing each new
"function" (regardless of size) by starting with a random landscape
from an 'average' (or really, a maximally distant) sequence is all
yours.

You haven't been able to explain the landscape of real novel systems
of function using the mechanism of RM/NS - not even close.  You
"explanations" rely completely upon your imagination.  

They rely of *evidence*. Real *evidence*. Not probability numbers
generated on the basis of pig-ignorant assumptions about mechanisms by
someone whose mind is so dichotomous he cannot think of *any* causal
feature other than an intelligent magical fairy.

What you do is
imagine what could have existed without any consideration of the odds
of your imagined scenarios actually existing in the real world.

I imagine what could have existed upon consideration of reasonable
precursors that actually *do* exist in *real* genomes*. You are the
one who has no evidence that the *only* possible alternative to pure
chance from scratch is a magical fairy. I am presenting a *causal*
explanation based on structures that *do* exist in *real* cells, not a
*causal* explanation based on a magic fairy that has no supporting
evidence whatsoever aside from false dichotomy in your fairy-besotted
brain. We *both* agree that complete random assembly from scratch is
not a viable mechanism. It is just that you seem unable to come up
with any causal explanation that can be tested while I can.

 You
yourself state that the likelihood of your imaginations reflecting
reality is "not computable".  Well Howard, that means that your
imagined scenarios are not scientific.  They are just-so fairytale
stories with no scientific value whatsoever.


Says the guy whose only possible alternative to pure chance is a
magical invisible fairy.

You're living in a fantasy world here.
This isn't science guess.  This is just-so story telling at its best -
or worst.

That the pattern of familial (i.e., descent with modification)
relationships actually exists has been statistically determined to the
point of being undeniable.

No one is arguing over the fact that NHPs exist.  They do.  

And they exist in families in which members show similar structures
and sequences but perform different *functions*. That is, the new
functions are *modifications* of old structures, not assembly anew
from scratch by a completely random process.

That has
absolutely nothing to say about the origin of the functional
differences arising via your proposed mechanism of RM/NS - nothing at
all.

So, you are saying that the similarity in sequence and structure of
the aldosterone and cortisol receptors, myoglobin and the hemoglobins,
alpha and beta hemoglobins (which form an IC system), and many other
proteins of relatively recent origin, is what? Pure chance? God's
little joke on us? These are functional differences that can *easily*
arise by RM/NS. In fact, as far as anyone can tell, *all* the
differences in sequence between humans and chimps (which went through
several intermediate species) can be *easily* attributable to RM/NS.
In fact, most could be and likely are due to chance alone. Only a few
differences in these two species *require* NS. How much more can be
accomplished and expected in the much longer time frames since, say,
the beta globins have arisen by modification of alpha globins (there
is consistent evidence of alpha's ancestry to beta).

[snip]

.

Relevant Pages

  • Re: whered Sean go?
    ... beneficial gap problem is overwhelming for the mechanism of RM/NS. ... is limited to <10% of all proteins. ... Multiprotein systems cannot, AFAICT, fit on your sequence landscape ... doesn't mean it is at a high level of functional complexity. ...
    (talk.origins)
  • Re: Howard theory - "cannot be computed"
    ... While scientific theories do have predictive power, ... mutations happen to hit upon a novel beneficial sequence "target". ... You don't understand the mainstream theory of RM/NS then. ... Otherwise, Hamming distance is undefined. ...
    (talk.origins)
  • Re: sequence points in subexpressions
    ... It may be hard to imagine how the C system ... The commas at left separate left-side sequence points, ... while OP's query *does* represent a defining corner- ...
    (comp.lang.c)
  • Re: infinity
    ... >>> Since 'sequence' implies that each item in the sequence is ... >>> followed by a unique 'next' item, can TO imagine what is the ... David Kastrup, Kriemhildstr. ...
    (sci.math)
  • Re: sequence points in subexpressions
    ... It may be hard to imagine how the C system ... There is a sequence point between the evaluation of i++ and the evaluation of i to its right, and it is the result of that i which is yielded by the comma operator and then has 1 added to it before being assigned to i. So, the sequence point of the comma operator is before the assignment side effect of the equals operator. ... while OP's query *does* represent a defining corner- ...
    (comp.lang.c)