More Recent Comments

Friday, August 31, 2007

Defining Irreducible Complexity


So many IDiots, so little time ....

Granville Sewell has posted a message on Uncommon Descent asking What if we DID find irreducibly complex biological features?. He writes,
In any debate on Intelligent Design, there is a question I have long wished to see posed to ID opponents: “If we DID discover some biological feature that was irreducibly complex, to your satisfication and to the satisfaction of all reasonable observers, would that justify the design inference?” (Of course, I believe we have found thousands of such features, but never mind that.)

If the answer is yes, we just haven’t found any such thing yet, then all the constantly-repeated philosophical arguments that “ID is not science” immediately fall. If the answer is no, then at least the lay observer will be able to understand what is going on here, that Darwinism is not grounded on empirical evidence but a philosophy.
Here's how Michael Behe defines irreducible complexity in Darwin's Black Box (p. 39).
By irreducibly complex I mean a single system composed of several well-matched, interacting parts that contribute to the basic function, wherein the removal of any one of the parts causes the system to cease functioning. An irreducibly complex system cannot be produced directly (that is, by continuously improving the initial function, which continues to work by the same mechanism) by slight, successive modifications of a precursor system, because any precursor to an irreducibly complex system that is missing a part is by definition nonfunctional.
There are many irreducibly complex systems in biology. One of my favorites is the citric acid cycle or Krebs cycle. This is a circular pathway of enzymes that oxidize acetate groups to two molecules of CO2.


If you remove any one of the enzymes then there is no cycle and it will be impossible to oxidize acetyl groups to CO2 and regenerate oxaloacetate. You cannot evolve a cycle for the complete net oxidation of acetate by starting with a more simple circular pathway then adding additional enzymes to improve the initial function; namely, the cyclic pathway of oxidation. Thus, by Behe's definition this is an irreducibly complex system whose function is to oxidize acetyl groups and regenerate the original precursor.

We have a damn good idea how the citric acid cycle evolved so the answer to Granville Sewell's original question is: no, the discovery of an irreducibly complex system does not justify the design inference. There are many ways of evolving irreducibly complex systems. This is the same answer that we've been giving for over ten years. Please try and keep up.

Now I have a question for the IDiots. If we can prove to your satisfaction that a particular system is irreducibly complex and demonstrate how it could easily have evolved, will you stop claiming that irreducibly complex systems can't evolve?

27 comments :

lee_merrill said...

"If we can prove to your satisfaction that a particular system is irreducibly complex and demonstrate how it could easily have evolved, will you stop claiming that irreducibly complex systems can't evolve?"

Ah, the battle joined! Behe's point is not so much about the required parts, as the required steps, which he makes plain in his reply to Miller and others:

"An irreducibly complex evolutionary pathway is one that contains one or more unselected steps (that is, one or more necessary-but-unselected mutations). The degree of irreducible complexity is the number of unselected steps in the pathway. ... If the improbability of the pathway exceeds the available probabilistic resources (roughly the number of organisms over the relevant time in the relevant phylogenetic branch) then Darwinism is deemed an unlikely explanation and intelligent design a likely one."

So then no, this is not problematic if a mousetrap can evolve step by successive selectable step.

Unknown said...

The creationist argument from irreducible complexity is tedious and, at this point, quite boring.

Creationist: "It's irreducibly complex, therefore it could not possibly have evolved!"
Scientist: "Well, there is at least one way that it could have evolved..."
Creationist: "But you have no proof that it happened exactly like that, so God must have done it!"

Torbjörn Larsson said...

Behe's point is not so much about the required parts, as the required steps

That is after moving the goalpost to two "improbable" steps to get function.

But both of these phenomena are different from the design idea. As Larry mentions the original subtractive definition overlaps with other theories and that goes for the newer step definition as well.

Granville Sewell isn't good at logics or science. (He is the one which invented "entropies" of macrostates (!) such as computer parts among other absurdities.) I guess Muller's definition of interlocking complexity and predicting it as an evolutionary feature in the beginning of the 20th century hasn't reached Sewell yet.

Of evolving interlocking complex mouse traps I like McDonald's best.

RBH said...

Lee_merrill wrote Ah, the battle joined! Behe's point is not so much about the required parts, as the required steps, which he makes plain in his reply to Miller and others:

Have you forgotten, lee, that structures that are irreducibly complex by Behe's structural definition (quoted by Larry above) evolved via pathways that are irreducibly complex by Behe's pathway definition that you quoted, doing so in Avida (Lenski, et al., Nature 423 (2003), 139-144).

Behe has never counted the number of unselected steps in any evolutionary pathway that led to an allegedly IC structure, nor have you. But I have done so in the Avida lineages, and not only do neutral mutations end up as parts of irreducibly complex structures, even mutations that are mildly deleterious on their first appearance do so. Both of Behe's notions of the unevolvability (or even the improbability) of irreducibly complex structures are blown away in that one study.

Must we repeat here the interminable avoidance of that fact you displayed on Infidels in this thread? I call readers' particular attention to this opening post and this post that summarize the data.

Anonymous said...

Excellent points as usual, but I'd like to ask whether you think the majority of the prominent IDers
A] genuinely believe their own claims have merit (and therefore it may theoretically be possible to reason with them); or
B] are simply liars and are only interested in winning over the general public, so of course are not interested in any kind of scientific dialogue but rather attempt to appeal to the ignorant directly.

I think most of them fit B better than A. If this is the case, aren't you just wasting your time with posts such as this one?

RBH said...

I'm going to add this post to the recommended readings. I rather like the summary I made there. :)

Anonymous said...

ISTM that a structure that can result from evolution can't be *irreducibly* complex as a matter of correct English, Behe's sloppy definition of "irreducible" aside. Thus, I'd much rather say likely evolutionary pathways have been shown for many systems that may have appeared to be irreducibly complex, rather than saying "irreducible" systems evolved from simpler precursors.

Unknown said...

Larry, I don't believe that we will ever find anything that is irreducibly complex, but I do think that it is an interesting question you sort of side-stepped. Behe's definition is obviously not a sufficient one for determining irreducible complexity; but if we really did have some way of determining that a system could not have evolved (proving a negative though...), what would it be evidence of?

I don't want anyone to get confused here, I'm not at all in support of intelligent design. I'm a biochemistry student not a nut-job. I'm just playing philosopher/Devil's Advocate.

lee_merrill said...

> "As Larry mentions the original subtractive definition overlaps with other theories and that goes for the newer step definition as well."

I’m not sure what you mean, but it overlaps with theories, I think the point still stands that multiple steps which are unselected with regard to the structure of interest, make that structure less likely to evolve. This is not a point that would seem to be disputable.

> RBH: "Have you forgotten, lee, that structures that are irreducibly complex by Behe's structural definition (quoted by Larry above) evolved via pathways that are irreducibly complex by Behe's pathway definition that you quoted, doing so in Avida (Lenski, et al., Nature 423 (2003), 139-144)."

Yet I’m sure you remember that the EQU function in the last scenario did not evolve, because of the number of unselected steps that would be required to reach it. And as far as your posts at Internet Infidels, I recommend in response, my replies.

But I'm not going to rehash all that here...

Larry Moran said...

Matt asks,

...if we really did have some way of determining that a system could not have evolved (proving a negative though...), what would it be evidence of?

First, by equating irreducibly complex with "cannot evolve" you have fallen into the trap set by the IDiots. They would like you to believe that whenever something is irreducibly complex it poses a problem for evolutionary theory.

My intent was to take their "scientific" definition of irreducibly complex and show that the conclusion was invalid. Irreducibly complex systems, such as bacterial flagella and the citric acid cycle can evolve from more simple systems that carry out different functions.

Now, on to your question ...

If it could be shown that something could not possibly have evolved then that would be evidence that there's something else going on that scientists don't know about. It would tell us that evolution is not sufficient to explain life.

Torbjörn Larsson said...
This comment has been removed by the author.
Torbjörn Larsson said...
This comment has been removed by the author.
Torbjörn Larsson said...

Aaargh! Maybe third time is the charm:

lee_merrill:

I’m not sure what you mean, but it overlaps with theories,

Sorry. RBH describes this better, as Behe's original structural definition (which is a subset of a scientific interlocking complexity definition) vs the one you discuss, the pathway definition.

I think the point still stands that multiple steps which are unselected with regard to the structure of interest, make that structure less likely to evolve.

IANAB, but AFAIU fixating a neutral or negative trait in a large but finite population has less probability than when selection is at work.

That a priori probability doesn't translate to a posteriori likelihood, to "less likely". That would be the usual texas sharpshooter fallacy and misunderstanding of basic probability theory of creationists. We can do better here and indeed, AFAIK this only means it takes more generations to find such fixated traits in large populations.

If we look at likelihoods for surviving populations they are AFAIU likely to have evolved traits of all types.

Torbjörn Larsson said...

lee_merrill:

I should add that the likelihood for different mechanisms and what is the null hypothesis in different situations seems to be part of what is currently discussed on other posts here, between "adaptionists" and other species of biologists. You are welcome to study a real scientific debate.

(Well, a blog "debate". The real debate in the corridors and on conferences is probably a lot more fierce. :-P)

lee_merrill said...

> AFAIU fixating a neutral or negative trait in a large but finite population has less probability than when selection is at work.

Yes, so then less likely to evolve?

> That a priori probability doesn't translate to a posteriori likelihood, to "less likely". That would be the usual texas sharpshooter fallacy ...

Actually, an a priori probability does translate into an after the fact priority, if the class of events of interest remains the same, and you got your a priori probability right.

Anonymous said...

Matt, you have a good point: Behe's definition of IC is confusing as standard English, which it is not.
Behe makes the two words together a single new technical term. The confusion still works after 11 years. People tend to think IC means 'unevolvable' even though the definition means coadapted parts, a normal result of evolution.

By the way Larry, your definition is too long. This first part is it:

"By irreducibly complex I mean a single system composed of several well-matched, interacting parts that contribute to the basic function, wherein the removal of any one of the parts causes the system to effectively cease functioning." [emphasis in original, which is Darwin's Black Box p 39]

The rest is Behe's deduction.

Lee, you have let the creationists mess with your mind. You probably even think Behe is a real biologist. As for quoting the time Behe ventured a completely different dfn of IC, he has repeatedly used the classical one since then.
And still never answered Miller .

The Dembskian words "probabilistic resources" have no merit, in part because these boys have no idea of the ways evolution can happen, and in part because they are constitutionally unable to apply statistics to biology sensibly.

The bottom line remains: IC is a normal outcome of evolution.




Pete Dunkelberg

Torbjörn Larsson said...

lee_merrill:

Actually, an a priori probability does translate into an after the fact priority, if the class of events of interest remains the same, and you got your a priori probability right.

No, since a likelihood is a conditional probability. You must specify your condition.

For example, if you ask if any verified evolutionary mechanism is likely to be observed, you get the answer yes (100 %). I covered that.

If you ask which evolutionary mechanisms are most likely to be observed, it depends on the history of populations, the current state of observations, the currents state of theory et cetera. It isn't an easy challenge to answer. Here is an exposition of 9 different mechanisms of speciation, and you can predict and observe (so test) their respective likelihood.

AFAIU in the genome it seems the mechanisms that Behe considers most improbable is the most likely to be observed and indeed the most observed. (Near neutral drift.)

Torbjörn Larsson said...

Sorry, my bad, of course Behe doesn't explicitly mention verified mechanisms, he tries to avoid known science results and theory. He considers pathways that AFAIU most often are explained by drift.

lee_merrill said...

"As for quoting the time Behe ventured a completely different dfn of IC, he has repeatedly used the classical one since then.
And still never answered Miller."

I posted a link to Behe's answer to Miller, though--and these definitions, one erases the other?

But if you all want to have a discussion in detail, then maybe start at thread over at Internet Infidels and point me to it.

Torbjörn Larsson said...

lee_merrill:


and these definitions, one erases the other?


They are independent.

(Which may reflect Behe's attempt of "scholarship". How did he handle the transition? Since they are falsified ideas which doesn't survive to be published even, we can't discuss them in term of hypotheses. But Behe must have gone through some motions.)

Anonymous said...

Lee,

The two definitions are not of the same thing at all. One (the real dfn of IC that Behe uses again and again) refers to the way something is. The other refers to how Behe thinks if got that way.

You didn't really post a link to Behe answering Miller. I don't think you groked the
Miller link I gave. Behe answers a strawman version of Miller. His answer to StrawMiller isn't so good either; what does he have against neutral mutations? But he ignores real Miller.

The important difference between Miller and Creationists like Behe is this: Miller is a scientist and is seriously trying to learn about nature works. Behe and friends just sit around making up arguments.

As this thread has died I'll probably make this point again elsewhere.

Pete Dunkelberg

Tone G said...

He says "I believe we have found thousands" as if he's trying to convince himslef. why not just say "there is thousands." there is no point in taking the argument seriously when (the vague) evidence: "thousands" (please can u tell me 20?) has the caveat of "i believe." sounds like he wants to believe and is trying tp convince himself.

Unknown said...

I think using counter-examples like CAC is good, but doesn't hit Behe's argument on the head. Saying IC systems cannot evolve is like telling someone to walk home in a straight line; then proclaiming, when they fail, that they must have appeared before you out of thin air. And that's 2D, with evolution we're talking about walking on a n-D fitness landscape.

IC assumes that nothing about the proteins (in the pathway) or the organism changes when the mutations we are interested occur. This is why scientists give counter-examples in lobsters for the clotting pathway, lobsters (and there clotting system) are different enough in other traits to sustain a simpler system.

-Rob

Unknown said...

I think using counter-examples like CAC is good, but doesn't hit Behe's argument on the head. Saying IC systems cannot evolve is like telling someone to walk home in a straight line; then proclaiming, when they fail, that they must have appeared before you out of thin air. And that's 2D, with evolution we're talking about walking on a n-D fitness landscape.

wrossite said...

Way late on this. Can Dr. Moran please tell me what this "damn good idea of how [the citric acid cycle] evoled" is? He claimed it, but I didn't see any specific model or experiment demonstrated. Thanks.

Larry Moran said...

Many bacteria don't have a functioning citric acid cycle. Instead, they have a forked pathway. One branch is a series of reductive reactions leading from oxaloacetate and acetyl-CoA to succinyl-CoA. (The left side of the citric acid cycle.) Succinyl-CoA is needed to make heme and other complex molecules in the cell.

The other fork corresponds to part of the right-hand side of the citric acid cycle. This oxidative pathway leads to formation of α-ketoglutarate, which is essential in amino acid synthesis.

It's reasonable to assume that these are the primitive pathways.

The "cycle" was formed with the evolution of the α-ketoglutarate dehydrogenase enzyme which links α-ketoglutarate and succinyl-CoA.

α-ketoglutarate dehydrogenase is clearly derived from pyruvate dehydrogenase so we have an excellent working model of the evolution of the citric acid cycle from two pathways that had different functions.

This example shows that irreducibly complex pathways can arise by evolution.

txpiper said...

Larry,

"The "cycle" was formed with the evolution of the α-ketoglutarate dehydrogenase enzyme which links α-ketoglutarate and succinyl-CoA."

What are the mechanisms involved in enzyme complexes evolving?