Irreducible Complexity

Subject: All Functions are Irreducibly Complex

Print | Individual message | Show original | Report this message | Find messages by this author

Discussions concerning the concept of irreducible complexity (IC) are
quite common and usually involve some confusion about the definition
of IC. The following question posed by “sds” sparked my interest.

“sds” <bcnorw@mindspring.com.leavethispartoff> wrote in message <news:bbcj98$t4d$1@slb6.atl.mindspring.net>…
> Do you believe is it possible to show that an object or mechanism has an
> intelligent origin? Take the mousetrap again (for which I understand you
> *don’t* find “irreducibly complex” to be a meaningful characterization).
> Can we show that a mousetrap probably could not exist without some
> intelligence to design it?

It seems to me that the concept of IC is quite helpful indeed. The
problem is that many, even Behe himself, seem to try to limit the
definition of IC to “very complex” systems of function in order to
show that IC systems cannot evolve.

As I see it all systems of function are IC. It is just that some
systems are more simple than other systems of function. There is a
spectrum of complexity, but all systems along this spectrum from
simple to more and more complex are all IC. In other words, not all
setups of a given number of parts or part types will be able to
perform a given function. The parts in any system of function can in
fact be altered, removed, or ordered in a different manner so that the
function of the system is completely destroyed. In fact, there are
vastly more non-functional potential arrangements of parts than there
are beneficially functional arrangements of parts in a particular
scenario.

Take, for example, Behe’s famous mousetrap IC illustration. Many try
to argue that a mousetrap is not IC since parts can be removed or
changed and it still can catch mice. That is not the issue. If you
change the mousetrap, it may still catch mice, but not in the same
way. The changed mousetrap is a different mousetrap that catches mice
in a different way. Certainly there are many different kinds of
mousetraps that can catch mice, some more effectively than others.
However, all of these mousetraps are dependent upon a certain number
of parts that are all arranged in a very specific way in order for
these parts to work together to catch mice (i.e., To perform their
function). Clearly there are a lot more arrangements of mousetrap
parts for any given type of mousetrap that would not catch mice at
all. All mousetraps can in fact be reduced or changed in a way that
would destroy their function completely. And, these potential
non-functional mousetraps are far more numerous than those
comparatively few arrangements than can actually perform the mouse
catching function.

Of course, it is theoretically possible to arrange several of these
working mousetraps in sequential order so that very small steps seem
to exist as one moves from one type of trap to the other. Obviously
then, it is NOT impossible for IC systems to evolve via function-based
selection mechanisms since such an evolutionary path need not
necessarily cross wide neutral gaps in function or non-function. The
problem is that these gaps are often wider than one might initially

think.

http://naturalselection.0catch.com/Files/irreduciblemousetrap.html

Even functions that are based on the workings of single proteins, such
as the enzymatic functions of lactase or nylonase, are IC in that
there are a limited number of parts that are required to give rise to
that particular type of function. For more simple functions, such as
these single-protein-based functions, there might be a much higher
ratio of sequences of a given length or smaller that would be able to
perform a given function, like the lactase or nylonase function. For
example, given a sequence of amino acids 1,000aa in size, there are
about 1 x 10e1300 different possible protein sequences. This is an
absolutely huge number of different possibilities. It is a 1 with
1,300 zeros following it. Out of all of these possibilities, how many
would have the lactase function? Certainly there would be many of
these sequences that would have the lactase function, but certainly
not all of them or even most of them. Perhaps the ratio would be as
high as 1 in a trillion? If the ratio where 1 in a trillion, that
means that any given functional lactase AA sequence would be
surrounded by an average of 1 trillion non-lactase sequences. If a
particular functioning lactase sequence is changed or “reduced” beyond
a certain point, it will no longer function at all, not even a little
bit. This is the definition of IC. The lactase function, even though
based in the AA sequence of a single protein, is IC. Of course,
compared to other systems of function, the lactase and nylonase single
protein enzymes are not all that complex since there are is a
relatively high percentage of potential lactase sequences as compared
with the total number of possible sequences out there. Because of
this, these functions are relatively simple, requiring a relatively
short stretch of DNA to code for their function. Other systems of
function require multiple proteins all working together
simultaneously. Much more DNA real estate is necessary.

Before thinking about more complex systems function, such as bacterial
motility, consider that even the evolution of the relatively simple
lactase function is quite difficult. Barry Hall demonstrated this in
several experiments where he deleted the lacZ genes in E. coli
bacteria to see if they would evolve the lactase function back again
using some other genetic sequence. And, they did evolve the lactase
function in just one or two generations. As it turned out, a single
point mutation to a completely different DNA sequence was able to
produce a selectively advantageous lactase function in a lactose
environment. Hall called this “evolved” sequence the ebg gene
(evolved beta galactosidase gene). But, he started wondering, “If
this worked for the deletion of the lacZ gene, what will happen if I
delete the ebg gene too?” So, Hall deleted the ebg and lacZ genes in
certain colonies of E. coli. What happened next is very interesting.
These double mutant E. coli colonies never evolved the lactase
function back again despite high population numbers, high mutation
rates, 4 million base pairs of DNA each, positive selection pressure,
and tens of thousands of generations.

http://naturalselection.0catch.com/Files/galactosidaseevolution.html

Now, why didn’t Hall’s double mutant E. coli colonies evolve the
relatively simple lactase function back again? Hall himself described
this colonies as having, “limited evolutionary potential.” What was
it that limited their ability to evolve the relatively simple lactase
function despite very positive benefits if they were to ever evolve
this helpful function?

It seems that neutral gaps existed between what was there and what was
needed. The genetic real estate of this huge population of E. coli
simply was not large enough to undergo the random walk across this
neutral gap in beneficial function despite being given thousands of
generations.

Obviously then, even such simple functions as the function of single
proteins are IC and this can and often does create difficulties for
mindless evolutionary processes. The problems only increase
(exponentially) as one moves up the spectrum of complex systems.

Sean
www.naturalselection.0catch.com

Discussions concerning the concept of irreducible complexity (IC) are
quite common and usually involve some confusion about the definition
of IC. The following question posed by “sds” sparked my interest.

“sds” <bcnorw@mindspring.com.leavethispartoff> wrote in message <news:bbcj98$t4d$1@slb6.atl.mindspring.net>…
> Do you believe is it possible to show that an object or mechanism has an
> intelligent origin? Take the mousetrap again (for which I understand you
> *don’t* find “irreducibly complex” to be a meaningful characterization).
> Can we show that a mousetrap probably could not exist without some
> intelligence to design it?

It seems to me that the concept of IC is quite helpful indeed. The
problem is that many, even Behe himself, seem to try to limit the
definition of IC to “very complex” systems of function in order to
show that IC systems cannot evolve.

As I see it all systems of function are IC. It is just that some
systems are more simple than other systems of function. There is a
spectrum of complexity, but all systems along this spectrum from
simple to more and more complex are all IC. In other words, not all
setups of a given number of parts or part types will be able to
perform a given function. The parts in any system of function can in
fact be altered, removed, or ordered in a different manner so that the
function of the system is completely destroyed. In fact, there are
vastly more non-functional potential arrangements of parts than there
are beneficially functional arrangements of parts in a particular
scenario.

Take, for example, Behe’s famous mousetrap IC illustration. Many try
to argue that a mousetrap is not IC since parts can be removed or
changed and it still can catch mice. That is not the issue. If you
change the mousetrap, it may still catch mice, but not in the same
way. The changed mousetrap is a different mousetrap that catches mice
in a different way. Certainly there are many different kinds of
mousetraps that can catch mice, some more effectively than others.
However, all of these mousetraps are dependent upon a certain number
of parts that are all arranged in a very specific way in order for
these parts to work together to catch mice (i.e., To perform their
function). Clearly there are a lot more arrangements of mousetrap
parts for any given type of mousetrap that would not catch mice at
all. All mousetraps can in fact be reduced or changed in a way that
would destroy their function completely. And, these potential
non-functional mousetraps are far more numerous than those
comparatively few arrangements than can actually perform the mouse
catching function.

Of course, it is theoretically possible to arrange several of these
working mousetraps in sequential order so that very small steps seem
to exist as one moves from one type of trap to the other. Obviously
then, it is NOT impossible for IC systems to evolve via function-based
selection mechanisms since such an evolutionary path need not
necessarily cross wide neutral gaps in function or non-function. The
problem is that these gaps are often wider than one might initially
think.

http://naturalselection.0catch.com/Files/irreduciblemousetrap.html

Even functions that are based on the workings of single proteins, such
as the enzymatic functions of lactase or nylonase, are IC in that
there are a limited number of parts that are required to give rise to
that particular type of function. For more simple functions, such as
these single-protein-based functions, there might be a much higher
ratio of sequences of a given length or smaller that would be able to
perform a given function, like the lactase or nylonase function. For
example, given a sequence of amino acids 1,000aa in size, there are
about 1 x 10e1300 different possible protein sequences. This is an
absolutely huge number of different possibilities. It is a 1 with
1,300 zeros following it. Out of all of these possibilities, how many
would have the lactase function? Certainly there would be many of
these sequences that would have the lactase function, but certainly
not all of them or even most of them. Perhaps the ratio would be as
high as 1 in a trillion? If the ratio where 1 in a trillion, that
means that any given functional lactase AA sequence would be
surrounded by an average of 1 trillion non-lactase sequences. If a
particular functioning lactase sequence is changed or “reduced” beyond
a certain point, it will no longer function at all, not even a little
bit. This is the definition of IC. The lactase function, even though
based in the AA sequence of a single protein, is IC. Of course,
compared to other systems of function, the lactase and nylonase single
protein enzymes are not all that complex since there are is a
relatively high percentage of potential lactase sequences as compared
with the total number of possible sequences out there. Because of
this, these functions are relatively simple, requiring a relatively
short stretch of DNA to code for their function. Other systems of
function require multiple proteins all working together
simultaneously. Much more DNA real estate is necessary.

Before thinking about more complex systems function, such as bacterial
motility, consider that even the evolution of the relatively simple
lactase function is quite difficult. Barry Hall demonstrated this in
several experiments where he deleted the lacZ genes in E. coli
bacteria to see if they would evolve the lactase function back again
using some other genetic sequence. And, they did evolve the lactase
function in just one or two generations. As it turned out, a single
point mutation to a completely different DNA sequence was able to
produce a selectively advantageous lactase function in a lactose
environment. Hall called this “evolved” sequence the ebg gene
(evolved beta galactosidase gene). But, he started wondering, “If
this worked for the deletion of the lacZ gene, what will happen if I
delete the ebg gene too?” So, Hall deleted the ebg and lacZ genes in
certain colonies of E. coli. What happened next is very interesting.
These double mutant E. coli colonies never evolved the lactase
function back again despite high population numbers, high mutation
rates, 4 million base pairs of DNA each, positive selection pressure,
and tens of thousands of generations.

http://naturalselection.0catch.com/Files/galactosidaseevolution.html

Now, why didn’t Hall’s double mutant E. coli colonies evolve the
relatively simple lactase function back again? Hall himself described
this colonies as having, “limited evolutionary potential.” What was
it that limited their ability to evolve the relatively simple lactase
function despite very positive benefits if they were to ever evolve
this helpful function?

It seems that neutral gaps existed between what was there and what was
needed. The genetic real estate of this huge population of E. coli
simply was not large enough to undergo the random walk across this
neutral gap in beneficial function despite being given thousands of
generations.

Obviously then, even such simple functions as the function of single
proteins are IC and this can and often does create difficulties for
mindless evolutionary processes. The problems only increase
(exponentially) as one moves up the spectrum of complex systems.

Sean
www.naturalselection.0catch.com

It seems to me that the concept of IC is quite helpful indeed. The
>problem is that many, even Behe himself, seem to try to limit the
>definition of IC to “very complex” systems of function in order to
>show that IC systems cannot evolve.

>As I see it all systems of function are IC. […]

So? We know very well that IC is not the least bit of an obstacle to
evolution.


Mark Isaak a@earthlink.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


Discussions concerning the concept of irreducible complexity (IC) are
> quite common and usually involve some confusion about the definition
> of IC. The following question posed by “sds” sparked my interest.

> “sds” <bcnorw@mindspring.com.leavethispartoff> wrote in message

<news:bbcj98$t4d$1@slb6.atl.mindspring.net>…

Behe relies on very complex systems because he realizes that an
argument as you use in the following paragraphs is too weak to hold up
under criticism.

As I see it, your argument depends on an unspoken assumption that
context (selective environment) stays fixed, but that is an assumption
that is not valid. In nature, there are few contexts so stable that
opportunities for improvement do not present themselves, and indeed in
those few cases, we can find ancient organisms who have been able to
sustain form and function as long as that context has been sustained.
Without selective context, evolution will happen only slowly in the
form of individual genetic arrangements and optimization.

But in an organism who has a great deal of “evolvability” –that is, a
systemic toolbox that is able to allow that organism entry into other
contexts, we should see appearence of function from existing parts
that are easily seen to be a result of co-option of existing pieces of
existing systems, OUT of their original context. In the case of a
mousetrap, the first step could only require a block of wood in a
completely different context. Then, nature sequentially rummages
through the vast “kitchen drawer” of biological tools and eventually
comes up with a mousetrap, as you’ve noted, in sequential steps.

At each step of parts generation, a new and different context (ie,
selective environment) was being answered to. A mousetrap was not
necessarily the item needed in predecessor contexts — however, the
majority of the parts, from the spring to the board — should have
existed in some early form before optimization. The predecessor parts
of the mousetrap, appeared not because of a need for a mousetrap, and
indeed if the mousetrap pieces were not already in existence, the
mousetrap would have never been assembled when the context for “need
mousetrap” arose.

From predecessor states of other existing systems (a pair of scissors,
a pen with a spring, a wire hanger), a mousetrap eventually is able to
be generated as the system’s “kitchen drawer” became more and more
populated with evolved tools generated from other contexts. The main
driving force for such experimentation and shuffling is not only born
out of the “messy” assembly experiments that we find in nature, but
also the need FOR the appearence of the context demanding “mouse
trap”. This isn’t mere supposition, but is born out of observed
instance.

Nylonase, as you pointed out, is one such example of an utilization of
existing systems to process a substance in a new context (that of
man-made nylon abundance). Yet, we hear about the successes. There are
plenty of examples of life NOT being able to adapt. For instance,
these bacteria that generated nylonase were probably not able to
quickly generate an enzymatic system that can feed off of granite. In
order to evolve a function, an organism must have predecessor units
from other contexts. Additionally, the right context must be present.
Contexts and opportunities (such as a rich diet in nylon fiber) must
exist in which to allow selection.

Therefore, the problem for Intelligent Design is NOT trying to show
“irreducible complexity” because, despite what Behe says, “IC” can be
generated easily by adaptation and refinement of an already not-IC
system.

The true problem for ID is to show that any biological system is
unable to adapt to new contexts utilizing existing parts (except there
are already instances of this occuring in nature). Beyond denying
existing evidence, ID could show a certain “neatness” of biological
function. Except we find that biological systems are full of noise and
re-arrangements, as if these systems were constantly attempting to
produce a large playing field of adaptations in order for something to
select for advantage.

ID’ers must show that biological systems are perfect machines that
cannot move from one context to another context simply by adapting —
by evolutionary process as outlined — their existing systems to other
contexts. Arguing that yeast cannot adapt to space conditions (as an
hypothetical example) is no argument, however, since evolution does
not require such great leaps. Evolution only requires reasonable
context jumps — nylon instead of granite. The largest problem for ID
is, however, that they’d be arguing against observation.

It is the evolutionists, not the ID’ers who are faced with
insurmountable difficulties. Evolutionists (especially darwinists) must
demonstrate that these functional adaptations can be discovered by
random, non-directed processes in spite of the fact that mathematical
analysis has shown a random search strategy to be highly inefficient.
For every functional system, the number of non-functional alternatives
is nearly infinite. To find these isolated “islands of function” from
within a “sea of noise” would be truly miraculous. Intelligent guidance
reduces the number of non-functional alternatives and directs the system
to optimum function is a much shorter time. To say that natural
selection is the operative equivalent of intelligent guidance and
accomplishes this task is blindly optimistic and not grounded in reality.

n article <3EDB7443.4020@charliewagner.com>, Charlie Wagner wrote:
> It is the evolutionists, not the ID’ers who are faced with
> insurmountable difficulties. Evolutionists (especially darwinists) must
> demonstrate that these functional adaptations can be discovered by
> random, non-directed processes in spite of the fact that mathematical
> analysis has shown a random search strategy to be highly inefficient.

What does random search have to do with evolution, Charlie?

> For every functional system, the number of non-functional alternatives
> is nearly infinite. To find these isolated “islands of function” from
> within a “sea of noise” would be truly miraculous.

Not miraculous at all, as genetic algorithms clearly and reliably demonstrate.

> Intelligent guidance reduces the number of non-functional
> alternatives and directs the system to optimum function is a much
> shorter time.

There is that “intelligent guidance” phrase you like so much. It’s a pity
that you don’t know what it is. Why not just call it “obfuscatonium”?

> To say that natural selection is the operative equivalent of
> intelligent guidance and accomplishes this task is blindly optimistic
> and not grounded in reality.

>>Discussions concerning the concept of irreducible complexity (IC) are
> >>quite common and usually involve some confusion about the definition
> >>of IC. The following question posed by “sds” sparked my interest.

> >>”sds” <bcnorw@mindspring.com.leavethispartoff> wrote in message

<news:bbcj98$t4d$1@slb6.atl.mindspring.net>…

– Hide quoted text –

– Show quoted text –

> >>>Do you believe is it possible to show that an object or mechanism has
an
> >>>intelligent origin? Take the mousetrap again (for which I understand
you
> >>>*don’t* find “irreducibly complex” to be a meaningful
characterization).
> >>>Can we show that a mousetrap probably could not exist without some
> >>>intelligence to design it?

> >>It seems to me that the concept of IC is quite helpful indeed. The
> >>problem is that many, even Behe himself, seem to try to limit the
> >>definition of IC to “very complex” systems of function in order to
> >>show that IC systems cannot evolve.

> > Behe relies on very complex systems because he realizes that an
> > argument as you use in the following paragraphs is too weak to hold up
> > under criticism.

> > As I see it, your argument depends on an unspoken assumption that
> > context (selective environment) stays fixed, but that is an assumption
> > that is not valid. In nature, there are few contexts so stable that
> > opportunities for improvement do not present themselves, and indeed in
> > those few cases, we can find ancient organisms who have been able to
> > sustain form and function as long as that context has been sustained.
> > Without selective context, evolution will happen only slowly in the
> > form of individual genetic arrangements and optimization.

> > But in an organism who has a great deal of “evolvability” –that is, a
> > systemic toolbox that is able to allow that organism entry into other
> > contexts, we should see appearence of function from existing parts
> > that are easily seen to be a result of co-option of existing pieces of
> > existing systems, OUT of their original context. In the case of a
> > mousetrap, the first step could only require a block of wood in a
> > completely different context. Then, nature sequentially rummages
> > through the vast “kitchen drawer” of biological tools and eventually
> > comes up with a mousetrap, as you’ve noted, in sequential steps.

> > At each step of parts generation, a new and different context (ie,
> > selective environment) was being answered to. A mousetrap was not
> > necessarily the item needed in predecessor contexts — however, the
> > majority of the parts, from the spring to the board — should have
> > existed in some early form before optimization. The predecessor parts
> > of the mousetrap, appeared not because of a need for a mousetrap, and
> > indeed if the mousetrap pieces were not already in existence, the
> > mousetrap would have never been assembled when the context for “need
> > mousetrap” arose.

> > From predecessor states of other existing systems (a pair of scissors,
> > a pen with a spring, a wire hanger), a mousetrap eventually is able to
> > be generated as the system’s “kitchen drawer” became more and more
> > populated with evolved tools generated from other contexts. The main
> > driving force for such experimentation and shuffling is not only born
> > out of the “messy” assembly experiments that we find in nature, but
> > also the need FOR the appearence of the context demanding “mouse
> > trap”. This isn’t mere supposition, but is born out of observed
> > instance.

> > Nylonase, as you pointed out, is one such example of an utilization of
> > existing systems to process a substance in a new context (that of
> > man-made nylon abundance). Yet, we hear about the successes. There are
> > plenty of examples of life NOT being able to adapt. For instance,
> > these bacteria that generated nylonase were probably not able to
> > quickly generate an enzymatic system that can feed off of granite. In
> > order to evolve a function, an organism must have predecessor units
> > from other contexts. Additionally, the right context must be present.
> > Contexts and opportunities (such as a rich diet in nylon fiber) must
> > exist in which to allow selection.

> > Therefore, the problem for Intelligent Design is NOT trying to show
> > “irreducible complexity” because, despite what Behe says, “IC” can be
> > generated easily by adaptation and refinement of an already not-IC
> > system.

> > The true problem for ID is to show that any biological system is
> > unable to adapt to new contexts utilizing existing parts (except there
> > are already instances of this occuring in nature). Beyond denying
> > existing evidence, ID could show a certain “neatness” of biological
> > function. Except we find that biological systems are full of noise and
> > re-arrangements, as if these systems were constantly attempting to
> > produce a large playing field of adaptations in order for something to
> > select for advantage.

> > ID’ers must show that biological systems are perfect machines that
> > cannot move from one context to another context simply by adapting —
> > by evolutionary process as outlined — their existing systems to other
> > contexts. Arguing that yeast cannot adapt to space conditions (as an
> > hypothetical example) is no argument, however, since evolution does
> > not require such great leaps. Evolution only requires reasonable
> > context jumps — nylon instead of granite. The largest problem for ID
> > is, however, that they’d be arguing against observation.

> It is the evolutionists, not the ID’ers who are faced with
> insurmountable difficulties. Evolutionists (especially darwinists) must
> demonstrate that these functional adaptations can be discovered by
> random, non-directed processes in spite of the fact that mathematical
> analysis has shown a random search strategy to be highly inefficient.

How come? The question is not one that would apply to the theory of
evolution. Natural selection is not random, in fact, it is decidedly
non-random. I am afraid your mathematical analysis was unnecessary.

> For every functional system, the number of non-functional alternatives
> is nearly infinite. To find these isolated “islands of function” from
> within a “sea of noise” would be truly miraculous. Intelligent guidance
> reduces the number of non-functional alternatives and directs the system
> to optimum function is a much shorter time. To say that natural
> selection is the operative equivalent of intelligent guidance and
> accomplishes this task is blindly optimistic and not grounded in reality.

Demolish that strawman Charlie! Just whack at it. Hope it makes you feel
better.

Frank