Re: Sean Pitman: definitions wanted

Seanpit wrote:
> hersheyhv wrote:
>
> < snip >
>
> > > The ToE really does say that the gaps do not increase significantly
> > > with increasing functional complexity. I say they do.
> >
> > *You* assert that they do, not me. That means you are not arguing
> > against the mechanisms that I (and other biologists) would propose.
>
> Exactly . . . I'm arguing against what you and other evolutionists
> actually propose.

No you don't. You are arguing against a mechanism that we say would
NOT work, but are claiming and asserting without evidence that this
mechanism (the one that would NOT work, namely crossing large neutral
gaps) is the only possible mechanism for evolution. It is that
assumption that is being denied.

> I say that what you really do propose really won't
> work because you do not understand that the neutral gaps really do grow
> with increasing levels of functional complexity.

You have presented precisely zero evidence that the size of neutral
gaps between selectable *functions* (meaning the number of independent
mutational events until a selectable *function*) *must* increase with
the size of the sequence or the number of amino acids involved in
structure (which are the only measures of "functional complexity" that
you present). My specific claim is that, for many of your examples,
the *function* (at least at a selectable level in environments that
lack that function) can arise in one or a few mutational steps
*regardless* of the size of the sequence or the number of amino acids
involved in structure (what you call "fairly specified" sites).

> That is not a strawman attack.

Of course it is. You are arguing against a mechanism that we
scientists do not hold and do so by making an unwarranted and
unevidence assumption that we scientists do not agree to. That is, you
are arguing against a mechanism we do not think happens and avoid
discussing the false assumption (that the size of gaps between
functional and selectable intermediates *must* increase with either
total sequence size or the fraction thereof that you claim is "fairly
specified").

> You may say that it is wrong, but you
> cannot say that I'm misrepresenting your position. I'm not. I'm just
> saying that your position is wrong - your real position. There is no
> misrepresentation of your position in my argument. I'm arguing against
> what you're actually saying, not some strawman representation.

You are arguing about the reality of a mechanism that we do NOT think
*must* happen. We don't think it *must* happen because we think your
assumption about the size of gaps (number of mutational events required
between functionally selectable intermediates) is completely bogus and
unsupported by any evidence whatsoever. We accuse you of engaging in
numerology rather than science. There is no way that your argument,
which is against a position that scientists do NOT hold based on
assumptions that scientists do NOT make, can be anything but a strawman
argument. A strawman argument is *defined* as an argument against a
position that your opponents do NOT make. And I firmly and
unequivocably reject the idea that evolution occurs by the mechanism of
"crossing large neutral gaps requiring many mutational events between
selectable *functions*" in favor of the mechanism that scientists do
hold, namely "descent with modification".

> > You are arguing specifically against an argument that I am not making;
> > that is, against the strawman argument that I am not making.
>
> Your argument, your real argument, is that there are no significant
> neutral gaps with increasing functional complexity. I am arguing
> against this real notion of yours.

Then you must present the evidence that your assumption that the size
of gaps (the number of mutational steps between selectable *functions*)
*must* *always* increase exponentially based on what you falsely call
*functional* complexity, but which really is either total sequence size
of some protein or system or the fraction thereof that is involved in
maintaining structure. And again, the *total* number of either measure
you make of the falsely named "functional complexity" makes the absurd
assumption that these systems evolve from scratch *each and every time*
a sequence appears and that there can be no functional or selectable
intermediate between a sequence that has none of the 'fairly specified'
sites and a sequence that has all of them. You make the absurd
assumption that evolution cannot borrow existing features nor evolve
functions by adding on or modifying functions to a pre-existing
sequence that has most of the final structure. You make the absurd
assumption that one cannot cross a gap by a series of steps, but only
by one large leap. That is what your math and the numbers you keep
using tells us (until pressed, when you then get all huffy and claim
that when you say that flagellar rotary motility *function* requires
crossing a gap of 10,000 steps by a random process with no intermediate
function you really are talking about 50 steps).

> I say that there are increasing
> neutral gaps with functional complexity. You say that I have no
> evidence for this. This is fine, but you cannot say that I'm arguing
> against a strawman version of your position. That's clearly not true.

The fact that you are arguing against a position based on a false
assumption that no one agrees with says that you are arguing against a
strawman. You will be arguing against a strawman regardless of what
you think until you come to grips with and present the evidence that
your unsupported assertion is correct, despite what we can actually
show (namely that the size of the gaps, the number of mutational steps
between selectable functions, does NOT increase exponentially with
either total sequence or the fraction of same that is 'fairly
specified').

> > You do
> > not present evidence that the only way that a particular end point can
> > be reached is by crossing your hypothetical large gaps.
>
> That's a different argument than suggesting a strawman attack against
> your position.

No its not. Again, a strawman attack is not an attack against *my*
position. It is an attack against some position that I do NOT hold and
the false claim that this position that I do not HOLD is my real
position. Do you understand what a "strawman" is? You do seem to
realize that you are arguing against a position that I do NOT hold
(because of the falseness of your unwarranted assumptions), yet you
seem to think that arguing against that position *that I do NOT hold*
is not a strawman argument, but is somehow an argument against the
position that I *do* hold (which is that your measure is bogus
numerology, that your assumption of increasing gaps as a function of
sequence size or a fraction thereof are unevidenced assumptions that
are actually falsified by evidence that you yourself has presented, and
that your numerology has no relationship to how evolution is thought to
have actually worked).

> You can argue that I don't have evidence to back myself
> up against your real position, but you cannot argue that I'm attacking
> a strawman version of your position. Understand?

It is precisely because you don't have evidence to back up your
assumptions and that I think your assumptions are false that leads me
to describe your argument as an argument against a strawman version of
evolution that no one really thinks happens (at least very often and
certainly not with gaps of 10 gazillion or whatever).

> > Instead you
> > keep *asserting without evidence* that *if* such gaps must be crossed,
> > then evolution would be so unlikely as to be essentially impossible.
>
> And you keep arguing that *if* such gaps did not exist that evolution
> wouldn't be a problem.

Well, we do have evidence that *when* such large gaps do not exist,
evolution by random mutation and natural selection don't have any
problem finding a solution. And we have *no* evidence that such large
gaps ever needed to be crossed in a single multi-step neutral process.
And, for even *functions* that you regard as "high level", such as
rotary motility, even you recognize that gaps no larger than 50 steps
are required and I point out that one can even, in principle, get the
*function* of rotary motility from precursor systems that have
different *functions* in as little as a single (or a few) mutational
steps. So given that we *know* that when the gaps are small (in terms
of number of mutational steps) between useful or potentially useful
*functions*, given that we can easily envision short gaps for even
*functions* you regard as complex (thus belying your assumption about
the relationship between *functional complexity* and the number of
mutational steps required to reach it), and given your *inability* to
present any evidence whatsoever that your numerological claims about
evolution working by starting in some random sequence in total sequence
space and reaching a specified teleological goal or even the smaller
'gaps' you handwave when pressed have any real-world reality, it is
*your* responsibility to demonstrate that there is some *function* that
requires crossing a large selectively neutral gap from a putative
functional intermediate or precursor. If you can't do that, your
assertion about increasing gaps is worthless, which makes you argument
against evolution occurring that way worthless.

> It is just that you have no evidence to support
> yourself beyond low levels of functional complexity.

You have repeatedly failed to describe "functional complexity" in a way
that makes any sense. Again, your descriptions of "functional
complexity" really is either of total sequence size in a system or a
fraction thereof (of 'fairly specified' sites) and has nothing to do
with any specified *function*. When you have *specified* an example
of what you mean, like rotational motility, I have demonstrated that
getting that *function* does not, in principle or necessarily, require
crossing a large neutral gap.

> Of course, you
> think you do have evidence and that I have none. That's a different
> issue than arguing for strawman attacks.

Not when you present an argument that I do NOT hold based on
assumptions I do NOT agree with. If you were to actually argue for the
idea that the size of gaps *must* increase with "functional complexity"
and present your evidence for that, that would not be a strawman
argument. But to assume the truth of your idea about exponential
increases in gap sizes, and then say that *because* your idea is true,
then evolution could not happen is not arguing against what evolution
says. It is arguing against something that would be true only *if*
your assumption were true.

> > And I agree.
>
> Well, there you go. If the gaps exist, evolution is impossible. If
> they don't exist, evolution is easy. There is no strawman here. This
> is the reality of the situation.

So, then. When are you going to present the evidence that there is a
correlation between gap size (the number of mutational events between
selectable and functional states) and your measures of "functional
complexity". The first thing you might do is specify how you measure
*functional* complexity and why you think that measure is a measure of
*function*. All I have seen is a conflation of sequence size (or a
fraction thereof) and *function*. Such a measure specifically denies
the possibility that a sequence can have one function and change to a
different function without having to make drastic changes in sequence.

> > So you are making an argument I agree with but not
> > presenting the evidence that the mechanism you argue against is even
> > considered to be the way that evolution works.
> > It isn't.
>
> I'm not arguing against the mechanism Howard. The mechanism is random
> mutation and function-based selection.

No. The mechanism you argue against is random mutation and NO
selection for a large number of mutational events and that the number
of such neutral mutational events increases exponentially as a function
of something you call *functional* complexity. I.e., the mechanism you
argue against *presumes* the existence of large gaps. The math you use
presumes that the starting point is always a random sequence far away
(differing at many sites) from a teleologically specified set of
sequences and that there is no selection until that teleologically
specified set is reached.

> Everyone agrees with that - even
> me. I'm arguing against your real notion that neutral gaps do not
> significantly increase with increasing levels of functional complexity.
> That is your real notion. That notion is not some strawman
> representation of your position. That's the real thing.

You have a funny way of doing this. You do it by assuming large gaps
and arguing against the impossibility of evolution *if* large gaps
existed. You do not, as you should, do it by presenting evidence that
there really is a correlation between the size of these neutral gaps
and what you call 'functional' complexity. So get cracking. Start
presenting the actual evidence that it is impossible for one to produce
function because of large gaps that *must* exist. Do it for some real
system. All you have done is present numerology based on the false
assumption of increasing neutral gap sizes. Your argument falls apart
when you actually look closely at that assumption.

> Now, you may
> say I'm wrong in challenging this notion of yours, but you cannot say
> that I'm challenging some strawman notion of yours. I'm not. I think
> your actual position, your real man of evolution, is wrong.

Then you need, again, to be clarifying the evidence that supports your
assumption about the relationship between "functional complexity" and
"size of neutral gaps". Perhaps you could start with mathematically
describing the "level of functional complexity" of hemoglobin beta's
oxygen-binding function, cytochrome c's electron transport function,
and the eubacterial flagella's rotary motion function and show how
those *functions* cannot possibly have arisen by crossing a short gap
between, say, the oxygen-binding capacity of hemoglobin alpha,
cytochrome c1, or the chimeric formation of FliG. Or present the
*function* that *must* arise by starting with a precursor of your
choice that lacks the function and can only reach its present state by
a neutral walk through sequences that lack the current function until
it is reached by chance after, say, 50, or 10000 'fairly specified'
different site changes.

> > And that is
> > what makes your argument an argument against a strawman, an argument
> > against a mechanism that no one but you considers worth arguing against
> > and NOT the mechanism that evolutionary biologists propose.
>
> You need to talk to someone about what a strawman argument really is.
> You evidently don't understand the concept.

I will when you start making an argument against what evolution really
proposes rather than an argument against a mechanism that is based on
your unexamined and asserted assumption. Or until you actually start
focusing on presenting evidence to support your assertion. Otherwise,
you give every impression that you think you are arguing against a real
evolutionary argument when you appear to say that *because* (rather
than *if*) your assertion is true, evolution is impossible.

> > > > Yes. *If* the distances are too long, that particular function would be
> > > > highly unlikely to have evolved.
> > >
> > > Exactly - that is my argument and it really does counter the true
> > > position of the ToE. It is therefore not a misrepresentation of the
> > > ToE. It is not an attack against a strawman.
> >
> > Yes it is. You are "mistaken" (despite being repeatedly informed that
> > you are mistaken) when you say that the mechanism of a random walk
> > across large neutral gaps is the "true position of the ToE".
>
> I do not say that random walks across large gaps is the *true position*
> of evolution. This is clearly not the position of evolution and I
> never said it was. This is MY position - not the position of the ToE.
> Get it? The real position of evolution is that these large gaps do not
> exist. I am arguing against the true position of evolution, the
> position that says the gaps do not exist, by presenting my own counter
> position - the position that the gaps do exist.

No. You are claiming that evolution cannot occur because it cannot
cross these large neutral gaps. You never present evidence for or
against the assumption that the large neutral gaps exist. You merely
assert that they do. I don't have to assert, of course, that small
gaps do exist and they are crossable by known evolutionary mechanisms.
We both agree that small gaps do exist. And their existence is *not* a
function of the size of the protein or the number of fairly specified
aa sites in these proteins. The small gaps exist because there is a
precursor structure that does not need to be changed much to produce a
novel or modified function in the organism in which such an event
happens. There, of course, is no guarantee that such a precursor will
exist in all organisms, and in those in which they don't, the
evolutionary event won't happen.

> Now, you can say that my position is wrong, but it is not a strawman
> attack against some false position that does not really represent how
> evolutionists think evolution is suppose to work. You cannot use the
> strawman argument.
>
> Get someone you trust to help you understand this.

Then let's have a discussion about the real point, which is the falsity
of your working assumption about the relationship between gap size
(number of required mutational events between selectable functions) and
some measure of "functional complexity" that makes sense (the measures
you have been using don't make sense). Don't waste our time by saying
that *if* your assumption is true, evolution wouldn't happen. We agree
on that.

> > It most
> > certainly is not the true position of the ToE, which is 'descent by
> > modification' rather than 'random walks across large gaps'.
>
> Again, I never said that the ToE suggests random walks across large
> gaps - It doesn't. I'm suggesting that random walks across very large
> gaps is the only possible mechanism at higher levels of functional
> complexity. This is my position, not the position of the ToE. Come on
> now Howard - this concept is not at all hard to understand.

And that is precisely what you need to present evidence to support.
You need to present evidence that random walks across very large gaps
is the *only possible* mechanism to reach some structure or function.
You have utterly failed to do so. I have presented alternative
possibilites for evolution that do not require such large gaps *between
selectable functions* for the examples *you* have proposed. Either
there is something *impossible* about my suggestions that you have not
explained or your unevidenced assumption has no support from nature.

> > > > But I am not interested in functions
> > > > that evolved by unlikely mechanisms. I am interested in the likely
> > > > mechanism that produced the functions that did appear.
> > >
> > > You assume the truth of evolution a priori. You can't do this in an
> > > argument that is all about challenging this basic a priori assumption.
> > > That's a logical fallacy.
> >
> > I am saying that *if* evolution produced a particular *function*, it
> > did so by modification of a pre-existing *structure* or system(s) that
> > did not perform that *function* *rather than* by starting from a random
> > sequence and crossing a large neutral gap.
>
> Ah, so you are saying "if" just like I am. You are arguing against my
> notion of large gaps by saying that if the gaps weren't there between
> what is and what could be that evolution would do just fine. That's
> exactly what I think. The problem is that the odds that your "if" is
> correct are extremely remote this side of trillions upon trillions of
> years of average time.

You have no support for this. Again, I have been proposing mechanisms
that occur much more frequently than your 'trillions' of years.

> <snip>
>
> > > > OTOH, I have many examples of systems and
> > > > functions that even you admit *have* evolved by crossing short gaps.
> > >
> > > Only at low levels of functional complexity. You've got nothing that
> > > has evolved beyond very low levels where the function in question was
> > > not historically in the gene pool to begin with - not a single example.
> >
> > Then there is no system that evolved at what you call a "high level of
> > functional complexity". Indeed, I have never seen you present a
> > measure of *functional* complexity. You have only presented either
> > total aa size or the fraction of those that are involved in
> > *structure*, without any connection to a specific *function*.
>
> Function is intimately connected to structure. There is a minimum size
> and specificity requirement for all types of functions. Certain types
> of functions require a minimum size that is very small and/or a minimum
> specificity level that is very low.

How does this relate to the size of the gaps between selectable
*functions*?

> However, other types of functions
> require a minimum size that is very large along with a minimum
> specificity that is also very high. The flagellar motility function is
> one of these high-level functions.

Not if my idea about the creation of FliG as a chimeric protein is
correct. At least wrt the number of mutational events required to
generate the specified *function* of rotary motility.

> It requires many thousand residues
> at minimum where each residue is at a fairly high level of minimum
> specificity (say, 1 in 1e30 per 100aa). The overall specificity for
> all of the thousands of residues involved for the whole structure
> creates an enormously isolated island of this type of function in
> sequence space. No other island of function comes even remotely close.

This is based on the utterly irrelevant idea that the flagellar
function appeared by the mechanism you admit is non-evolutionary.
Again, I have proposed generating the motility *function* in a single
mutational event (given precursor systems that have different
functions). Even you, but only when pressed, agree that the size of
the gaps between functionally useful islands is NOT "many thousand
residues" but only 50-60 (although you haven't said how you calculated
those numbers). But now you are again saying that "no other island of
function comes even remotely close"? So what is it? 50-60?
Thousands? Not even remotely close? Or one mutational event linking
two subsystems that can function and exist even in bacteria without
flagella? *You* have to demonstrate that my idea is *impossible*.
*You* have to choose which numbers you really mean (there is a bit of a
gap between 50-60 and thousands, after all). *You* have to present the
evidence that the motility *function* *must require* crossing a huge
selectively neutral gap requiring many mutational events that have no
effect until all of them have occurred. [I should point out that I
don't know how you are going to do this.]

>
> > And you
> > blindly refuse to see that proteins have sub-functions/structures, such
> > as binding heme, that do not need to be reinvented, but can be
> > borrowed.
>
> I've told you over and over again that higher-level functions, to
> include the flagellar system of motility, do have sub-structure
> functions that can and do work independently of the overlying
> higher-level function. The fact that you continue to say that I say
> otherwise is a strawman representation of my actual position. It is
> you, not me, who does the strawman building around here.

Bull***. *You* keep presenting bogus numbers that you *know* are
irrelevant, like your "thousands" and only start admitting that
higher-level "functions" do have sub-structure functions when forced
to. Just like your switching between 50-60 and "thousands". I
strongly smell disingenuousness, but won't claim it is the reason quite
yet. BTW, the subfunctions of the rotateable pore of flagella and the
motor of flagella did not change at all when the *function* of rotary
motility was produced by forming the chimeric linker. The pore is just
as rotateable and the motor still causes movement at a distance. It is
just that now we have rotary motility because the two subsystems
(neither having active rotary motility) became linked.

> > > > What you have to do is demonstrate some system that *must* cross a gap
> > > > as large as you are proposing. All you have done is present numerology
> > > > which says that *if* you *must* start with a system x gazillion aa's
> > > > away from any structure that has a *specified* (aka, teleologic)
> > > > function, *then* it could not evolve. Even if I agreed with the
> > > > numerology (and basically I do), you have failed to demonstrate that
> > > > any *real* system *must* evolve that way.
> > >
> > > I've presented functions at very high levels, like the flagellum.
> >
> > "Flagellum" is not a *function*.
>
> Flagellar motility is a function. You clearly know what I'm saying
> here. Why act like its not clear to you?

Good. Then let's discuss how 'flagellar motility' arose. Did it arise
by starting at some random sequence thousands of specified sites away
from the current system? Did it arise by crossing a gap of 50-60
mutational events, all of which are neutral until all are present? Or
did it arise by the formation of a crude motility function by the
formation of a chimeric linker between pre-existing motor and
rotateable pore and protein-transport tube in, possibly, as few as one
mutational step? Which possibility has the most specific information
consistent with that possibility?
>
> <snip>
>
> > > You
> > > seem to think that because it is possible to get multi-character
> > > mutations to happen in one shot that it is easy to evolve at such
> > > levels. The problem you don't seem to understand, is that the odds of
> > > a "just so" multi-character mutation coming along are much much
> > > smaller, exponentially smaller, with the increasing size and
> > > specificity requirements of the needed multi-character mutation.
> >
> > Even if the pore had five times as many proteins and the motor had 10
> > times as many, how does forming a single chimeric protein to link the
> > two subsystems increase the size and specificity requirements for the
> > single mutational step required.
>
> It is the size and specificity of the changes needed to properly link
> the two systems together that's the problem. The greater the number
> and specificity of the changes required to create the link, the
> exponentially greater the odds of failure.
>
> > And I am quite satisfied that the
> > mutation need not be as specific that it requires a chimera only be
> > between aa X on one protein and aa Y on another. The ability to form
> > functional fusion proteins between FliF and FliG (the two that did form
> > differ by more than 100 aa's more being deleted from one than the
> > other) tells me that the requirements are more flexible than that.
>
> A requirement for specific changes to just 150bp at a specific location
> in the genome (coding for just 50 residues) would be an uncrossable gap
> this side of trillions of years of average time.

Do you know how frequently deletions, insertions, inversions and
translocations occur in bacteria and their associated episomes? If you
did, you would not say this. There are no *changes* to 150 bp. At
least not any that need to occur stepwise. The formation of a chimeric
protein is due to a single mutational event, not 150. And again, there
is no reason to think that the event needs to be at a *specific*
location, only that it produce the chimeric ends and enough (or little)
of the intervening to allow the two ends so created to act as a linker
between the motor and FliF. Again, it is entirely possible that the
original mutation produced a fusion FliFG that subsequently produced
the current separate FliF and FliG proteins. I say this because the
site of current requisite interaction between the two proteins can be
eliminated by deletion to form a functional fusion FliFG by
*spontaneous* mutation in far less than a trillion years. That implies
that the site of interaction in the present FliF and FliG may have been
part of the compact FliF protein initially, with selection for compact
protein structure due to the interaction actually having arisen well
before the initial fusion event produced a fusion FliFG. In this case,
the conversion of the fusion protein to two proteins occurred at a
different site such that the interaction (which is non-covalent) still
occurred.

Besides, how frequently is it *necessary* for a flagellar function to
evolve? The correct answer, at present, from the evidence of nature,
is twice: once in the lineages of eubacteria and independently in the
lineage of archae. All eubacterial flagella, AFAIK, are due to common
descent rather than independent evolution.

> < snip >
>
> > > > You seem to assume that only a
> > > > fully functional, at modern levels of function, can be useful and that
> > > > proteins can only have a single *function*.
> > >
> > > Not true. I'm looking at minimum selectability of the crudest form of
> > > a particular function.
> >
> > And assuming that the only way to reach it is to change, by a gazillion
> > independent selectively neutral steps, *all* the aa's that are even
> > remotely involved in that protein. All the while refusing to examine
> > the sequences of likely pre-existing ancestral proteins that lack the
> > specified *function* but have most of the *structure*.
>
> There's nothing remotely close enough to anything at the level of
> functional complexity required by function like flagellar motility.
> There just are no "close" steppingstones like you imagine there are.

Motor. Rotateable pore (with a protein tube or pilin attached). Both
exist in organisms today. Evolution does not even require that they be
a specific type of motor or rotateable pore. The fact is that the
second independent evolution of a rotary flagella did not use the same
type of motor nor the same type of rotary pore plus pillin. The fact
that all eubacterial flagella use the same type of motor (analogs still
exist independently) and pore plus whip is due to common descent and
not independent evolution.

> The subsystem structures of the flagellar motility system are far far
> away from the minimum structure needed to gain flagellar motility.

How is one mutational step "far, far away from the minimum structure
needed to gain flagellar motility"? The *function* of flagellar
motility.

> The
> number of changes and degree of specificity of those changes needed to
> link up pre-existing subsystems just so is enormous in terms of random
> walk distances/time.

One mutational event to gain *function* is not "enormous in terms of
random walk distances/time". Even if the mutation were idiosyncratic.

> > > > Neither of these
> > > > assumptions are realistic or based on evidence.
> > >
> > > And, I make neither assumption. You are building a strawman to
> > > represent my position - yet again. You are the strawman builder Howard
> > > - not me.
> >
> > Your math tells a different story, Sean.
>
> No, it doesn't. My math is based on linear increases in average gap
> distances with increasing levels of functional complexity - not the
> changing of all the residues at random to get a system like flagellar
> motility. That is yet another one of your strawman versions of my
> position.

But you have not presented evidence to support this bogus idea of
linear increases nor any theory (except the bogus model of starting
from random sequence space maximally distant from a teleologic
structure island). And now the increases have become linear rather
than exponential! It is hard to keep up with all the wild movement
between thousands of changes to 50-60 to linear gaps to exponential
gaps. And not a single idea supported with evidence. But perhaps that
is why you handwave all over the place? Perhaps we should start with
something simple, like a discussion of how you measure "functional
complexity" in a way that is actually related to *function* rather than
total sequence size or a fraction thereof.

> > > > There are many
> > > > mutations that produce proteins with *quantitative* changes in function
> > > > that provide an easy path to optimization and most proteins have
> > > > secondary *functions*.
> > >
> > > Yep - but none of these facts help you beyond very low levels of
> > > functional complexity.
> >
> > Why do you think evolution needed to go beyond very low levels of
> > functional complexity (which you seem to *define* as a system that
> > requires crossing these hypothetical large gaps)? If you claim that
> > there is no system of "high levels of *functional* complexity" but only
> > systems that produced complexity by a series of "low level of
> > *functional* complexity" mutational steps, why would I disagree?
>
> But there are lots of systems at high-levels of functional complexity -
> like flagellar motility. Such high-level systems are never evolved -
> ever.

Actually, it evolved or appeared twice, using different substructures
(that had analogous function however). It is merely your assertion
that it cannot be evolved. Besides, it is still true that one of the
IC proteins of the current flagellar function has been deleted in
bacteria and, under selective pressure to evolve motility, a second
(pre-existing, naturally) protein has evolved (been modified) to
perform the lost function. I have no doubt that I could arrange
conditions where I delete the amino terminal end of FliG (knocking out
motility function) and then could select for deletions that form a
fusion FliFG protein that would restore function. And I have no doubt
that I could use a plasmid with IS elements or even homolgy no longer
than 6-7 nt to generate a fusion product between FliF and FliG's
carboxy terminal end that would restore function. And be able to do so
in far less than a trillion years.

> Not even proposed steps in their evolution are ever evolved -
> ever. You haven't got a single example - not one. Only low-level
> systems can be evolved in a reasonable amount of average time.

What is the difference between level of *functional complexity* and the
number of mutational steps between functional intermediates? If there
is no difference, then all you are saying is that *if* a system must
cross a large neutral gap (requiring many neutral mutational steps),
then it cannot evolve in a reasonable time and subsequently
*assserting* that there are actual examples of such systems. That is,
you are obfuscating the difference between what would happen *if* the
flagellar (or whatever function you call complex) function requires
crossing a large gap between functional sequences involving many
neutral mutational steps and what happened *because* the flagellar
function *must* cross such large gaps. You have yet to present
evidence that flagellar function *must* cross such large gaps.
Handwaving "thousands" of sites won't do. Neither will even the 50-60
you handwave when forced. You have to specifically present evidence
that it is *impossible* to produce any chimeric FliG-like protein that
could link a motor to a rotateable pore (with a pilin or tube attached)
that could produce a selectable flagellar motility *function* where
none existed.

> These
> are the only examples you have - low-level systems of function evolving
> (like lactase or nylonase, which require no more than a few hundred
> fairly specified residues at best).

I continue to reject the idea that the number of 'fairly specified
residues' is relevant to the mechanism by which evolution allows
acquisition of new or modified *function*.
>
> > > > > > This is obvious when you actually ask how the *functions* of your
> > > > > > example systems actually did arise (as in lactase *function* from ebg)
> > > > > > or could arise (as in flagellar *function* or cytochrome c *function*).
> > > > >
> > > > > Do you know why the function of lactase actually evolved successfully
> > > > > in Hall's E. coli? - or why it was possible for the nylonase function
> > > > > to evolve in other bacteria exposed to nylon over time? Because, both
> > > > > of these functions require no more than a few hundred fairly specified
> > > > > residues at minimum.
> > > >
> > > > No. Both those functions arose because they could produce the new
> > > > *function* from a *structure* that didn't have that *function* by one
> > > > or a few mutational steps.
> > >
> > > Yes - and why were these functions only a few mutations steps away from
> > > what already existed?
> >
> > Such "Why?" questions are the wrong questions to ask in science.
>
> LOL - you've gotta be kidding! ; )

> > It
> > simply is a fact that there sometimes exist novel functions that are
> > only a few steps away from what already exists.
>
> Yes, and this "sometimes" is a lot more common at lower levels of
> functional complexity than it is at higher levels. At higher levels,
> this "sometimes" gets exponentially less and less until it becomes, for
> all practical purposes, "never".

And, to repeat, there is no "higher level" if you define 'higher level'
as those levels that *had to* evolve by crossing large gaps requiring
many neutral mutational steps. You have yet to present a single
example of a system that *must* have evolved by crossing such large
gaps. I say that the size of the gaps that must be crossed between
selectively useful steps does NOT increase at any point in the
evolution of any system that organisms actually have. The gaps that
are crossed always have involved no more than a few mutational steps
between functional intermediates. You keep going from what would be
the case *if* such gaps exist to stating it in terms of "this system
cannot evolve *because* such gaps exist". Remember, Sean, you have
already agreed that you are really saying *if* such gaps exist, not
*because* such gaps exist in order not to be producing a strawman
argument that assumes what you are trying to "prove" (actually
demonstrate as being consistent with the evidence) or is merely an
exercise in numerology. But that is just like your saying that you
don't really mean "thousands" of changes inbetween useful
intermediates, but only 50-60.

If you don't keep saying *if* when you talk about these hypothetical
large gaps, I will keep reminding you that that is what you need to
demonstrate.

> < snip >
>
>
> > > Because of the relatively high ratio that exists
> > > at these lower levels. That's why. What are the odds that such narrow
> > > gaps will exist at higher levels? Exponentially lower - that's the
> > > problem. The odds of narrow gaps get more and more remote at higher
> > > and higher levels because of the exponential decreasing ratios.
> >
> > You seem to be assuming both that larger gaps exist and that they
> > *must* be crossed by an increasing number of steps. Larger gaps may
> > exist, but I have no evidence that any system that does appear had to
> > cross such large gaps.
>
> Higher-level systems do not just "appear" without ID. That's the
> problem.

Actually that is an unevidenced assertion. And basing your entire
argument on an unevidenced assertion is a problem.

> You have no examples of higher-level evolution making any
> higher-level function "appear".

Since you seem to define 'high-level function' as those hypothetical
functions that require crossing large neutral gaps requiring many
neutral mutational steps, this is rather circular and does rather
assume that such high-level functions actually exist. I haven't seen
you actually present anything but assertions that such high-level
functions actually exist.

> You assume that evolution did make
> such systems appear in the past, but this is an a priori assumption
> which you cannot make in this discussion.

Nope. I simply don't see any evidence that high-level functions (that
is, functions that require a mechanism involving a random walk of many
mutational steps) actually exist rather than merely be proposed as a
hypothetical for the purpose of numerology. We both agree that
low-level (that is, functions that require no more than a few
mutational steps between selectable intermediates) can and have
evolved. But I cannot see any connection between *function* and your
measures of "functional complexity", which seems to be nothing but
sequence size or a fraction thereof).

> > I hear you asserting this again and again, but
> > I certainly have not seen any evidence that such large gaps *must* be
> > crossed to produce the systems that do exist.
>
> If the gaps are there, and they are, how were they crossed without ID?
> If you say that the "large gaps may exist" then you are forced to
> assume either ID or an extraordinary streak of luck for random mutation
> and function-based selection. What is the most tenable position?

Such large gaps certainly may exist and, in theory, do exist. But that
does not mean that any system that *actually* exists *must* have been
produced by crossing such hypothetical neutral multi-mutational gaps.
Certainly not if much shorter pathways to *function* exist, even in
principle, and certainly when supported by evidence, such as the
evidence I have presented for the rotary motility function and the
chimeric formation of FliG. That mechanism is much more likely than a
mechanism that involves a crossing involving thousands (or even 50)
mutational steps.

> <snip>
>
> > > > > What does this mean? It means that the ratio of
> > > > > nylonases and lactases in sequence space is relatively high.
> >
> > No it doesn't. It means that *function* may not be all that unusual in
> > searchable sequence space
>
> LOL - that's exactly what I just said. We are talking about the
> density of sequences with certain types of "functions" in sequence
> space.

There is a differnce between *function* and "certain types of
functions". The former makes no claim to specify *function*
beforehand. The latter does. And nylonase is just as complex as most
digestive enzymes.

> Lower-level functions have a much higher density of sequences
> in sequence space that carry a minimum degree of that type of function
> as compared to higher-level functions.

Again, it seems that the only definition of "lower-level" functions are
those functions that have been shown to be produced by a few mutational
steps. It doesn't seem to have any correlation with the size or
specificity of the protein for substrate. Again, how do you actually
determine what functions are "lower-level" and which are
"higher-level"? What is the level of complexity of the beta globin of
hemoglobin? How do I calculate level of *functional* complexity? Or
are these terms nothing but your disingenous claim that anything that
can be shown to be evolvable is therefore "lower-level" and any
function that you think you can get away with claiming is unevolvable
is "higher-level". It certainly appears that way from here.

> > (remember that much of total sequence space
> > is not searchable, cells favoring compact proteins' persistence).
>
> That only adds to my argument. Limitations on what can be searched in
> sequence space makes it the average time involved to find beneficial
> sequences at higher level of functional complexity skyrocket.

The exact opposite. The smaller total sequence space is, the more
likely that the search will find functional moieties rather than
useless non-compact moieties, since functional moieties are
concentrated among compact proteins. However, since most functions
actually arise by modification of pre-existing *structurally related*
sequences that are also functional (often with related functionality),
real searches tend to be much shorter than even under this assumption.
This is reflected in the obvious bias toward protein families in living
organisms.
>
> > But
> > the evolution of specific proteins depends on the existence of the
> > appropriate source sequences that are not far from the novel function's
> > sequence.
>
> Exactly - but what happens when higher-level functions are not at all
> close to anything else in the gene pool?

The hypothetical "higher-level" (if by that you mean a hypothetical
function that hypothetically *must* be reached by a long random walk
involving many mutational steps) function wouldn't be present. Now all
you have to do is find a higher-level (as defined above) *function*
that actually does exist. That is, you have to come up with a function
that *must* have involved said long series of neutral mutational
events. You haven't done that yet. The examples you give simply are
assertions that they *must* have involved such a long series of steps.
Yet I have rather easily presented evolutionary 'short-cuts' that don't
require such a long walk. And done so by looking at the actual
evidence of the structure of the proteins and the functions of subparts
rather than by bogus numerology involving size of sequence or a
fraction thereof.

> <snip>
>
> > > > No single step in any *real* proposed model of the evolution of the
> > > > flagella requires anywhere near 10,000 residues.
> > >
> > > Read what I wrote again Howard. I've explained this to you over and
> > > over again, but you just don't seem to get it. The average gap between
> > > the flagellar system of motility and the nearest steppingstone island
> > > is *NOT* 10,000 residues wide. The flagellar system requires at least
> > > 10,000 fairly specified residues to achieve minimum selectable
> > > function, but this is NOT the gap. I repeat - this number does NOT
> > > represent the gap. Get it?
> >
> > Then why the f**k do you keep trotting out that bogus number.
>
> It's not a bogus number. This 10,000-residue minimum is a very good
> estimate of the minimum size requirement for the function of flagellar
> motility.

Of course, as far as the number of mutational steps required for
crossing any steppingstone island, it is as bogus and irrelevant as can
be.

> > You are
> > still confusing *structure* and *function*.
>
> Function is dependent on structure Howard. You simply cannot get the
> function of flagellar motility with significantly fewer than 10,000
> fairly specified residues.

Of course you can. But the *size* of the total number of proteins
present in modern flagella is not an indicator of the difficulty of
evolving a structure that has the rotary motility function. It is an
utterly irrelevant number. Even wrt to the size of the many
steppingstone gaps.

> > Since the *function* of
> > motility is acheived in a single step (specifically the one that links
> > the motAB function to the rotateable pore with whip), the number of
> > 10000 is a bogus strawman number which is completely irrelevant to the
> > evolution of the *function*.
>
> Not true. Even if you can get the function of flagellar motility with
> a single mutation, the minimum size and specificity requirement would
> still be about 10,000 fairly specified residues. That requirement does
> not change regardless of how the structure was actually formed.

SFW? The important number for evolution is the number of mutational
steps required between selectable functions. If the *function* of
flagellar motility can arise by a single mutational step, that means
that flagella do not meet your requirement for high-level functional
complexity.
>
> > And *you* did say "functions like
> > flagellar motility, which require some 10,000 fairly specified residues
> > at minimum".
>
> Flagellar motility does require at least 10,000 fairly specified
> residues to be in place at minimum. A lot of genetic real estate is
> required, at minimum, to code for the flagellar system of motility.

That would only be relevant if one thought that these aa's arose from a
random sequence by a neutral walk.
>
> > So it is hardly surprising that I might think that you
> > were claiming that the *function* of a flagella requires, at minimum,
> > 10000 fairly specified residues".
>
> The function of flagellar motility does require, at minimum, 10,000
> fairly specified residues. It may not require, however, 10,000
> residues changes to achieve given a subsystem starting point within the
> genome. These are two different concepts Howard. They are not the
> same thing.
>
> I've already explained this many times. I'm simply amazed that you
> find such concepts so difficult to grasp.

Then stick to the parts that are relevant to your argument rather than
b.s. about 10000 fairly specified residues that has no relevance
whatsoever.

> > > The actual gap between a subsystem steppingstone function and the
> > > flagellar function may be more like 50 or 60 residue differences wide.
> >
> > *If* the size of the relevant gap needed to produce flagellar
> > *function* (presumably rotary motility) is 50-60, then *that* is the
> > number you should use, not keep trotting out the bogus number of
> > 10,000.
>
> Not true. It is because of the minimum size requirement of at least
> 10,000 fairly specified residues that the gaps get as large as 50 or 60
> fairly specified residue differences. The gaps at lower levels, where
> functions only require a few hundred fairly specified residues at
> minimum, may only be 5 or 6 fairly specified residues differences wide.
> Make sense?

No. Again, there are only a limited number of evolutionary steps that
have the effect of altering *function*. None of these changes in
*function*, AFAICT, involve even gaps *requiring* 50 mutational steps.
But reaching a selectable *function* is not the same as optimizing that
*function* after a change has produced the selectable *function*. Most
of your 10,000 fairly specified positions are involved in optimizing
function rather than producing it. If, as I propose, a single
mutational event *produced* a selectable rotary motility function, that
in no way means that that function was anywhere near being optimized.
There could easily have been 50 or more additional changes that would
have a quantitative effect on the efficiency of rotary motility. But
random point mutation and selection is an excellent mechanism for
optimizing efficiency of a system with selectable function.

> > I am not the one trotting out that bogus number, Sean. I am
> > using the number you keep using and *claiming* is the number needed for
> > flagellar *function*.
>
> It's not a bogus number. You just interpret it in a bogus way.

You present it in a bogus way. As if it had some relevance to the
evolution of these systems. It doesn't. And your 50-60 number,
although more relevant, seems to be nothing but handwaving.

Again, what *you* have to demonstrate is that *some* step in the
evolution of *some* system *must* have *required* a random walk
involving a large number of mutational steps to produce *any*
selectable function that would reasonable intermediate in the evolution
of a currently existing system. If there are no such steps, then there
is no problem for evolution. Yet you keep asserting that your
hypothetical *if* there are such gaps really means *because* there are
such gaps.

All I have to do is say that *when* there is no such large gap, it is
clear that the known mechanisms of evolution can produce the observed
feature. And I see no evidence that there is any theoretical or other
reason why the size of the individual gaps between intermediates gets
larger with any measure of size or function. And no evidence that
there are features that *require* crossing such large gaps. So you can
hypothesize about the consequences of such large gaps all you want. As
far as I can tell, they don't exist except as hypothetical wishes on
your part.

> > I say that flagellar *function* can be potentially obtained by one
> > mutational event, specifically the formation of the chimeric hybrid
> > protein we call FliG.
>
> That's fine, but this one mutational event will still have to meet the
> minimum 10,000-residue requirement for minimum flagellar motility.

It does. Most of the flagellar complex is present after this mutational
step, and I have already pointed out that one would not have to even
make many changes in either the pore proteins or motor proteins, since
their functions really have not changed. [The duplication and
specialization of the flagellins is probably subsequent to the
evolution of motility rather than prior, so a lot of those aa's in the
current flagella would be subsequent to the evolution of flagellar
*function* rather than prior.]

> > So now the argument is between your hand-waving
> > number of 50-60 that appears out of nowhere and my number of one that
> > comes from an analysis of how the subsystems of a bacterial flagella
> > are put together and interact.
>
> Your own proposal of crossing the gaps involve at least 50 or 60
> residue changes to achieve success for most of proposed steps in
> flagellar evolution.

Not really. In fact, changes in or modification of selectable
*function* would probably only require a few mutational steps. And it
is number mutational steps that are important, not the number of aa's
that are changed in that step. Of course, once you have a selectable
*function*, mutation and selection would optimize sequence.
>
> > > I've said this over and over again. Why the strawman over and over
> > > again?
> >
> > Because *you* keep saying "functions like flagellar motility, which
> > require some 10,000 fairly specified residues at minimum", which
> > certainly does seem to imply that you think that that is the number you
> > mean. It seems that you only toss out other hand-waving numbers like
> > 50-60 (for some unspecified step in the formation of flagella) or 5-6
> > (for cytochrome c) when the bigger bogus number (which relates to
> > *structure* and not *function*) is challenged.
>
> You need to understand the difference between minimum structural
> requirements for certain types of functions and the minimum gap sizes
> involved. This will make a huge difference to your understanding of
> this problem.

Not really. The number is still irrelevant. It would only relevant
*if* one presumed that all those sites arose via a single random walk
from a random sequence that lacked all of them. That clearly is not
the case and even you admit it. Knowing that number tells me
absolutely nothing about the size of any *functional* gap (the number
of mutational changes or steps needed to change a function or modify it
so that it becomes primary) nor anything about the many sites that
evolved via optimization of a *function*. It tells me nothing about
the ease or difficulty of evolving any single step (or the most
difficult step). So basically, it tells me nothing of interest.

For example, I would expect considerable homology between the different
flagellins, with all of them sharing many of the functionally relevant
sites. But I certainly would not expect this to be any indication at
all about the difficulty of evolving that *similarity*. I certainly
would not think that it arose by each flagellin starting from scratch
to evolve that sequence.

The number is utterly useless as an indicator of anything of relevance
to the existence or not of some specific gap that *must* be crossed by
many selectively neutral mutation events.

As to the number 50-60, that seems to be nothing but a WAG. WAGs won't
do, Sean.

> > > Yes, a single mutational event that involved multiple character
> > > differences = very low odds of success this side of trillions of years
> > > of average time.
> >
> > Thus it is impossible to produce a fusion protein of FliF and FliG
> > because the amino acids at the amino end of FliG are necessary for FliG
> > *function*?
>
> It is not impossible, just very unlikely.

Yet, in less than a week or month, actual real scientists generated not
one but two different such fusion mutations in the small populations
(relative to the numbers in nature) they were able to grow in a lab.
That hardly seems to be unlikely when one considers geological
timeframes. You seem to have a very strange definition of unlikely.

> < snip >
>
> > > Function is largely dependent upon structure. There is nothing
> > > structurally or functionally homologous enough, not even close, to the
> > > flagellar system of motility.
> >
> > Evidence? All the whip proteins are modifications of whip proteins.
> > The motor proteins are clearly related to other motor systems.
>
> It's the way they are put together. Nothing is remotely homologous to
> the overall system. There are subsystem homologies, to be sure, but
> none of these comes close to the overall system. At best the overall
> system shares no more than 50 or 60% homology with any one subsystem.
> Even if it shared 95 or 99% homology with a subsystem, it would not be
> good enough.

Hell, Sean. You are talking about a system that evolved several
billion years ago. 50-60% homology with other systems, given the fact
that many aa sites are selectively neutral is damn good. I would feel
grateful to get any significant homology. We aren't talking about
something that evolved last week. There is such a thing as neutral
drift. And I have no idea what you mean when you say that there are
subsystem homologies, but none that comes close to the overall system.
If the overall system is due to the fusion of two independently evolved
subsystems, why should there be homology outside the subsystem from
which it came?
>
> < snip >
>
> Sean Pitman
> www.DetectingDesign.com

.