An early IC prediction
Søren K
I remember once finding a mention of a biologist in the first half of
the last century who predicted that evolution would sometimes make
systems that seemed impossible to develop stepwise. It was mentioned as
a counterpoint to Behes idea of irreducible complexity, which is
basically the same idea, but the other way around.
Do any of you know more about this. I need a citation or at least the
name of the biologist for a debate I'm having elsewhere, but I cannot
find it right now.
Thanks
Soren
James Laurenson
Five seconds at the archive produces this:
http://www.talkorigins.org/faqs/behe.html
>From your post it sounds like you have your chronology mixed up a bit,
people don't make counter-points 50 odd years before the first person
makes a point (and don't get all Paley on me).
And it isn't one person who is refuting Behe, it's the entire
scientific community. There are a host of articles demonstrating the
flaws in his ideas.
Søren K
James Laurenson wrote:
> >From your post it sounds like you have your chronology mixed up a bit,
> people don't make counter-points 50 odd years before the first person
> makes a point (and don't get all Paley on me).
Obviously it was a contemporary writer who brought up the point. And it
is a counter of sorts to the IC claim. If a scientist more than 50 year
argued that evolution must lead to IC (he didn't call it that), then
ID's case is somewhat weakened.
>
> And it isn't one person who is refuting Behe, it's the entire
> scientific community. There are a host of articles demonstrating the
> flaws in his ideas.
Yes I know. But I am looking for a citation for a specific prediction
from the start of the former century (I think it was in the 1930's),
that looks a lot like IC, but with the conclusion that ET predicts IC.
chris.h...@gmail.com
The name you might be looking for is Herman Muller.
Here's the detail, reworked from a post I wrote at an on-line web
forum.
Herman Muller, in 1918, indicated that an expected result of
evolutionary processes was the development of what he called
"interlocking complexity". Muller was one of the great geneticists of
the twentieth century. He went on to win a Nobel in 1946 for work
in mutations.
<http://nobelprize.org/nobel_prizes/medicine/laureates/1946/index.html>
Muller's definition of "interlocking complexity" is exactly the same
as the definition of "irreducible complexity" -- a system of mutually
independent parts that requires all those parts to be present for the
system to work. However, Muller's claim is that this is an EXPECTED
result of evolution. Behe took the same definition, and claimed it
was IMPOSSIBLE as a result of evolution.
The reason for the difference is basically that Muller was
using evolution; and Behe was using a weird strawman of his own
devising. Behe describes evolution as working by the gradual addition
of parts, one by one. Muller, however, describes evolution as working
by gradual modifications of parts. Muller's description is the more
accurate. New proteins don't get added to systems particularly often;
the vast majority of evolution is small modifications to proteins,
to alter their amino acid sequence and hence their chemistry. Behe
neglects this entirely; and hence omits the vast majority of
evolutionary change.
The paper is "Genetic Variablity, Twin Hybrids and Constant Hybrids, in
a Case of Balanced Lethal Factors", by Hermann J Muller, in "Genetics",
Vol 3, No 5, Sept 1918, pp 422-499. You can read a scan of the paper
online at <http://www.genetics.org/content/vol3/issue5/index.shtml>
Here is a relevant extract, from pages 463-464 of the article:
... Most present day animals are the result of a long process
of evolution, in which at least thousands of mutations must
have taken place. Each new mutant in turn must have derived
its survival value from the effect which it produced upon the
"reaction system" that had been brought into being by the many
previously formed factors in cooperation; thus a complicated
machine was gradually built up whose effective working was
dependent upon the interlocking action of very numerous
elementary parts or factors, and many of the characters are
factors which, when new, where originally merely an asset finally
become necessary because other necessary characters and factors
had subsequently become changed so as to be dependent on the
former. It must result, in consequence, that a dropping out of,
or even a slight change in any one of these parts is very likely
to disturb fatally the whole machinery; ...
The evolution of the blood clotting system can be used as
an example to show how this works. It's pertinent, as Behe
gives this as an example of IC. A more complete account
of how blood clotting cascades can arise is given in The
Evolution of Vertebrate Blood Clotting, by Ken Miller, online at
<http://www.millerandlevine.com/km/evol/DI/clot/Clotting.html>.
I'll paraphrase his section "Introducing Complexity", which gives
a simple example of just the process Muller described in 1918.
First, the core of blood clotting is a "clot-maker" protein, which has
a kind of "sticky" section in the middle of the protein, and this is
normally covered by smaller chains of amino acids. Another protein (a
"protease") can "activate" the clotting process by clipping off the
covering chains of the clot-maker, exposing the central section. The
clot-maker proteins then bind together into a solid mass that makes
up the clot.
However, the full process in humans, or any vertebrates, is much more
intricate, involving a whole chain of proteins, each of which acts to
"activate" the next in a whole chain of activations, ending up with
activation of the clot-maker. The claim of irreducible complexity is
basically the claim that removal of any protein from the chain would
make the whole process fail.
Here then is an example of how such "irreducible" mutual dependencies
can arise by evolution, illustrating the same process described by
Muller in 1918. Miller's article goes into more detail and covers
stages that I omit in this shorter account. I am simply focused here
on showing the evolution of mutual dependencies, or interlocking
complexity.
(A) Start with a system consisting simply of two proteins; the
clot-maker and the protease. The protease is "activated" by contact
with tissue proteins - as would happen when there is a break in a
blood vessel. The activated protease is then able to activate the
clot-maker, and the clot is formed.
(B) Now have a gene duplication for the protease. This is a reasonably
common process in evolution; an entire section of the genome gets
doubled; so that now there are two genes, both producing the same
protease protein. There is no difference to the working of blood
clotting; as all the proteins involved are the same.
(C) Now have a small modification to one of the duplicated genes. There
are now two slightly different forms of the protease. Call them
protease-A and protease-B. Either one would manage fine for blood
clotting. In that sense, the system of three proteins is no longer
irreducible; it has redundancy.
(D) Now suppose that there are mutations to protease-A which give
it a capacity to activate protease-B. That is, both proteins get
activated at the break in a vessel by contact with tissue proteins;
but protease-B gets additional activation from the activated
protease-A. This kind of additional activation can have some selective
benefits, in speeding up the response of the whole system.
(E) Finally, now that protease-B is activated by protease-A, it no
longer depends on activation from the tissue proteins, and further
modifications can reduce this activation pathway. This makes the
whole system "irreducible" again, because all three proteins are now
required for clotting.
The fundamental point here is that an irreducibly complex core
of three proteins can arise from a simpler system by evolutionary
changes. Behe's argument depends on a strawman of evolution. Behe
ignores the role of modifications to proteins, and bases his argument
simply on the problem with getting a system by adding parts one by one.
Miller's article goes into a lot more detail, describing how the
initial two protein system can form, and giving examples of known
organisms with functional blood clotting systems that do indeed
represent intermediate stages in this kind of development.
Muller did not know the details of blood clotting cascades, but
he had already in 1918 correctly anticipated the processes by which
a system composed of many mutually interdependent parts would be
expected to arise by evolution.
Cheers -- Chris Ho-Stuart
TomS
<1158840759.8...@m73g2000cwd.googlegroups.com>,
=?iso-8859-1?B?U/hyZW4gSw==?= stated..."
See the reply by Chris Ho-Stuart for more details about Muller.
But there were others who brought up an idea remarkably similar to
"irreducible complexity". Georges Cuvier, in the early 1800s, had a
concept of "correlation of parts", which he was rather famous for.
What has interested me is that there were a few students of "natural
history" of the 1700s who used the same idea to argue, not against
evolution, but against the idea of development of the embryo. Take
a look at the website that I refer to in my .sig for an exposition of
their anti-development arguments, comparing them to more recent
anti-evolution arguments.
So far, the earliest reference to this sort of thing that I've heard of
is from the scientist of the mid-1600s, Pierre Gassendi:
"... the works of Nature are not like the works of art which
are made only by progressing from one part to another, since,
unlike an artisan, Nature together with her instruments
resides right within the material and can act on all the parts
at once just as well as on a single one; ..."
Gassendi, quoted in volume 2, page 811:
Howard B. Adelmann
Marcello Malpighi and the Evolution of Embryology
Cornell U. Press, 1966
Which seems to be arguing that, rather than being an argument
for design, it is an argument for natural origins.
--
---Tom S. <http://talkreason.org/articles/chickegg.cfm>
"... have a clear idea of what you should expect if your hypothesis is correct,
and what you should observe if your hypothesis is wrong ... If you cannot do
this, then this is an indicator that your hypothesis may be too vague."
RV Clarke & JE Eck: Crime Analysis for Problem Solvers - step 20
chris.h...@gmail.com
chris.h...@gmail.com wrote:
> Muller's definition of "interlocking complexity" is exactly the same
> as the definition of "irreducible complexity" -- a system of mutually
> independent parts that requires all those parts to be present for the
Urk. Typo. That should read "system of mutually INTERdependent parts"
CreateThis
wrote:
I believe this is mentioned in Michael Shermer's book "Why Darwin
Matters", but I'll have to look when I get home.
CT
asbes...@gmail.com
Søren K wrote:
> Hi
>
> I remember once finding a mention of a biologist in the first half of
> the last century who predicted that evolution would sometimes make
> systems that seemed impossible to develop stepwise.
There's also Gould's work on exaptation, which may fit this bill. His
basic point was that if we only look at structures as being evolved for
their *current* function, there are some cases which we can't explain
through gradual evolution. The classic example is the wing: if a wing
is just for flying, then 5% of a wing doesn't do anyone any good, so
the wing could never have evolved.
The solution, of course, is that the gradual evolution could have been
helped along by a different function: in this case the usefulness of 5%
of a feather flap as a good thermoregulator. So structures that look
like the could never have evolved in one adaptive landscape can
actually evolve in another, and then switch over from one landscape to
another.
If you're interested, I actually wrote my master's thesis on exaptation
in genetic algorithms (computer programs), but it's fairly technical.
Matt Silberstein
<Soren.K...@gmail.com> in
<1158840759.8...@m73g2000cwd.googlegroups.com> wrote:
Muller.
--
Matt Silberstein
Do something today about the Darfur Genocide
http://www.beawitness.org
http://www.darfurgenocide.org
http://www.savedarfur.org
"Darfur: A Genocide We can Stop"
Desertphile
Mark Isaak
> ... I am looking for a citation for a specific prediction
> from the start of the former century (I think it was in the 1930's),
> that looks a lot like IC, but with the conclusion that ET predicts IC.
Hermann J. Muller, 1918 and 1930. Full references in
/indexcc/CB/CB200.html.
--
Mark Isaak eciton (at) earthlink (dot) net
"Voice or no voice, the people can always be brought to the bidding of
the leaders. That is easy. All you have to do is tell them they are
being attacked, and denounce the pacifists for lack of patriotism and
exposing the country to danger." -- Hermann Goering
Victor Eijkhout
This is very interesting.
Victor.
Bob Casanova
talk.origins, posted by chris.h...@gmail.com:
Thanks! This one goes into my "Saved Posts" folder.
--
Bob C.
"Evidence confirming an observation is
evidence that the observation is wrong."
- McNameless
Hopeless Monster
> On Thu, 21 Sep 2006 06:27:07 -0700, Søren K wrote:
>
>> ... I am looking for a citation for a specific prediction
>> from the start of the former century (I think it was in the 1930's),
>> that looks a lot like IC, but with the conclusion that ET predicts IC.
>
> Hermann J. Muller, 1918 and 1930. Full references in
> /indexcc/CB/CB200.html.
>
I have been trying to find those papers on internet. I've found the
paper published in 1918(thanks to Chris Ho-Stuart) but not the one that
was published 1939. I haven't got access to libraries where that paper
could be found.
It would be nice if someone who has access to that paper could provide
some actual quotes from it. How exactly Muller defines "interlocking
complexity"(does he even use that word?) and so on.
Søren K
replied before now.
Thanks all for your answers - it was right on!
Søren K
nominated it for POTM, so I wont have to
Cheers
Søren
Ron O
chris.h...@gmail.com wrote:
> Soren K wrote:
> > Hi
> >
> > I remember once finding a mention of a biologist in the first half of
> > the last century who predicted that evolution would sometimes make
> > systems that seemed impossible to develop stepwise. It was mentioned as
> > a counterpoint to Behes idea of irreducible complexity, which is
> > basically the same idea, but the other way around.
> >
> > Do any of you know more about this. I need a citation or at least the
> > name of the biologist for a debate I'm having elsewhere, but I cannot
> > find it right now.
> >
> > Thanks
> >
> > Soren
>
> The name you might be looking for is Herman Muller.
>
SNIP:
Good post.
One nit. It turned out that Behe didn't know what he was talking about
for IC systems. One of the early definitions that he put forward was
the interlocking parts type system where if you take a part away it
stops doing its normal function. Those systems turned out to not be
the type of IC systems that Behe needed. What he needed was a system
that could not have evolved by natural known mechanisms. His simple
definition failed. Behe's type of IC system might still exist in
nature and have interdependent parts, but even Behe knows that he needs
something else to determine if such a system is his type of IC. So
Behe is still searching for a definition of IC that would be useful for
the inferences that he is trying to make. He tried to make a big deal
about "well matched" parts, but he couldn't tell anyone how to measure
it or how well matched parts had to be to satisfy him, or how well
matched the parts of his system had to be to produce selectable
function. The last thing that I saw from Behe on this topic was that
the more parts a system had the more IC it is. You no longer need just
interdependent parts, but a lot of them. The problem is that Behe
can't tell anyone how many parts are enough to satisfy his inference.
It looks like IC has degenerated to the old creationist probability
argument that depends on probabilities that no one can calculate. If
IC had amounted to anything we would have heard about it in Dover, and
Behe would have published by now in a real science journal. Behe
testified that he could not test and that he was not testing his ID
notions (unless he excludes himself from his stated state of ID at the
moment). Behe testified that no one had tested ID and no one was
testing ID, but that you could test ID like in science fiction, if not
in the real world. IC is still at the stage where they are trying to
figure out if they have anything worth making inferences from. They
haven't even gotten to the stage where they can test if they can infer
ID from any such system. All they can claim is that "if" their type of
IC systems exist in nature, that they "might" be able to make some type
of design inferences from such systems. Since they don't know what an
IC system is, they haven't even gotten to the "might" part.
All the current ID notions are like this. ID failed in Dover because
it never got beyond the wishful thinking stage. They haven't gotten to
the point where they can do any real science. A lot of neat scientific
ideas start out that way, but we don't teach the ones that haven't made
the grade, yet, for the simple reason that the vast majority of such
notions fail once we are able to test them. Not only that, but ID has
been a consistent failure throughout the history of science. Who makes
babies? Who pulls the sun and moon across the sky? Who causes
disease? Who made the world 10,000 years ago? etc. The only thing
that separates those past ID failures from the current "Who made the
flagellum?" is the level of ignorance science was dealing with at the
time. ID has a 100% failure rate upon testing in science, so is there
any reason to teach it in the science class except as an example of how
not to do science?
Ron Okimoto