More Recent Comments

Thursday, December 06, 2007

Irreducible Complexity

Lately I've run across several different definitions of irreducible complexity [Utterly Stupid Quote of the Day]. Most people seem to think that irreducible complexity is defined as something that cannot evolve but that's not the original definition [IDiot Logic].

Here's what Michael Behe says on page 39 of Darwin's Black Box.
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. 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.
Behe then goes on to say that an irreducibly complex system cannot evolve by natural selection.

The definition is not a problem. By this definition there are many irreducibly complex systems in biology. For example, the bacterial flagella is a pretty good example as long as you relax the criteria a little bit. (Some of the minor parts can easily be removed without affecting the overall function.)

The problem is not with the definition, it's with the conclusion. Irreducibly complex systems can easily evolve. All that's required is for the simpler intermediates to have some function other than the one seen in the final completed structure. In the case of the bacterial flagella this simpler function was secretion of large molecules. The flagella evolved from a type III secretion system by just adding a few extra components.

Thus, as Behe says above, it was not produced directly by continuously improving the initial function. Instead, there were several intermediate functions (e.g., secretion) that preceded the shift to the final function we observe today. This is how irreducibly complex systems evolve.

The citric acid cycle is another example of an irreducibly complex system for which there's an easily understood evolutionary pathway. The circular pathway arose when the ends of a forked pathway were joined by evolution of a single enzyme [Defining Irreducible Complexity].

Irreducible complexity is a concept invented by Intelligent Design Creationists. You'd think they would at least make the effort to understand something they created!


11 comments :

Unknown said...

Don't forget "scaffolding" (at least the term I used, following from the arch analogy), where a part that is initially unnecessary but beneficial becomes essential as the system evolves. To go back to the bacteria flagellum, there are parts that are absent in some bacteria flagella, present but not necessary in others, and present but necessary in others. So you have some organisms that don't have it, some organisms were it is not an irreducibly complex part, and other organisms where it is an irreducibly complex part. That kind of contradicts Behe's entire argument.

Anonymous said...

Isn't another flaw in Behe's argument (that his definition of IC is a problem for evolution) that an IC system can be the result of the loss of a part from a previously non-IC system?

For example, an arch consisting of stone blocks shows IC as the removal of any block causes the arch to collapse. However, the arch can be constructed by successive modifications where an initial support is in place. The subsequent removal of the support gives an IC system that has evolved through successive modifications, one of which was the loss of a previously required part.

Loss of gene function (and loss of whole genes) is a well known and evidenced phenomenon due to deleterious mutations. Therefore the loss of a previously required, but subsequently redundant, part gives an IC system that can evolve through successive additions followed by a deletion.

Anonymous said...

The expression, "irreducible complexity" was invented by creationists, but the concept is an old one. Check the Wikipedia article "Irreducible complexity", where under the heading "Forerunners" there is brief mention made of several writers who used ideas very much like IC, hundreds of years before Behe.

Particularly interesting is the use of something like IC to defend the idea of biological preformation. Preformation seems rather like creationism applied to the origins of individuals, rather than "kinds".

Anonymous said...

As originally formulated, IC systems are non-controversial. They exist and can be reasonably determined by deletion analyses. The controversy, as Larry notes, is whether a biological IC system is evolvable. ICness refers to the state of the system *as it exists today*. The problem is that many anti-evolutionists accepted Behe's arguments that IC systems couldn't evolve and thus were reliable indicators of (and effectively synonymous with) design. The trouble, as we all know, is that the "show your work" step of the proof never materialized.

But it's not just others that made the mistake of conflating ID with unevolvability...

Behe himself made subsequent new definitions to 'IC' that took into account the pathways of the system in question and in so doing basically defined an IC system as one that is likely to be unevolvable. The trouble was that those were radically different formulations (call them IC-Mark I and IC-Mark II); ones that presupposed a historical route and thus were not independent assessments of the system as it stands today. And he never did demonstrate that any system was "IC" based on these newer criteria. Yet Behe was more than happy use the same term "IC system" interchangeably for all three of his definition classes. Thus he also added to the muddling of the concept. Worse still, there were added terms such as "IC cores" which refer to the "essential" ICness at the core of a complex system such as blood clotting. Thus Behe and other supporters would say that some components of blood clotting which, while essential but clearly evolvable, were not part of the "IC core" and thus didn't detract from his original argument.

Go figure. Some people keep beating dead horses, thinking that they'll eventually get a ride.

Wavefunction said...

"...because any precursor to an irreducibly complex system that is missing a part is by definition nonfunctional."

I think Behe is being disingenuous here. He assumes irreducible complexity here and then when he says "by definition" he means "by definition of irreducible complexity". And of course, we know that precursors to these systems are functional even when they are missing a part. So I don't understand why this statement of his is sensible.

Torbjörn Larsson said...

However, the arch can be constructed by successive modifications where an initial support is in place.

That is the analogy of biological "scaffolding" I have come to expect, which explains why the key stone only appears to be essential at all times when in reality scaffold and key stone change roles.

under the heading "Forerunners" there is brief mention made of several writers who used ideas very much like IC

Looks like a good article on that point.

Most damning is that Muller's concept of interlocking complexity precedes Behe's in time and scope, which The Panda's Thumb has observed on several occasions. If Behe or any IDer were serious about ID as 'science', they would accept the scientific term. They would also have to show why the verification of Muller's prediction isn't part of evolution of course.

new definitions to 'IC' that took into account the pathways

That was an interesting analysis. I have distinguished between Behe's original deterministic claim ("impossible") and the subsequent probabilistic ("improbable" - but apparently not for his own example of malaria). But the discussion of history shows how much that Behe has come to surrender to science facts and theory.

All, really, as no one questions that evolution is contingent and any one pathway may in fact be a priori improbable in the face of all these possibilities. But the a posteriori likelihood that an evolutionary pathway is used has been found to be indistinguishable from 1. :-P

lee_merrill said...

> And he never did demonstrate that any system was "IC" based on these newer criteria.

That was most of his Black Box book, actually, demonstrating the implausibility of various steps that should be expected on all paths. Thus no need to presume a historical sequence, may it be noted.

> But the a posteriori likelihood that an evolutionary pathway is used has been found to be indistinguishable from 1.

Strange indeed, then, that the numbers seem to be coming from the IDers, and not from the naturalists, when they present their scenarios for evolution of flagella and mammalian blood clotting.

Anonymous said...

I am not familiar with any ID-er ever giving an estimate of probability that an "intelligent designer" would design something-or-other with "irreducible complexity".

Not to mention any justification for the value.

Let me give my estimate, just to have a point of departure, and then others can make corrections to it.

My first estimate for an unspecificed designer doing some unspecified thing at some time resulting in something that turned out to be the way things are: zero = one divided by infinity.

For a second estimate, I have an upper bound for the probability that ID would do it: Strictly less than the probability that natural causes would do it.

Does anyone have a better number?

lee_merrill said...

> Toms: For a second estimate, I have an upper bound for the probability that ID would do it: Strictly less than the probability that natural causes would do it.

In that case we reserve judgment, if based on what we know, we can compute a bound on the probability of nature generating a structure.

But if a good case can be made for a supernatural God (I think it can), then this God designing in the world is not like "unknown being X making Y in some unknown way Z." Design then would become more plausible, being more specific.

Here by the way is Behe's definition of IC in terms of steps: "If a system has to pass through one unselected step on the way to a particular improvement, then in a real evolutionary sense it is encountering irreducibility: two things have to happen (the mutation passing through the unselected step and the mutation that gives a selectable system) before natural selection can kick in again. If it has to pass through three or four unselected steps ... then in an evolutionary sense it is even more irreducibly complex. The focus is off of the 'parts' (whose number may stay the same even while the nature of the parts is changing) and re-directed toward 'steps.' "

"Envisioning IC in terms of selected or unselected steps thus puts the focus on the process of trying to build the system. A big advantage, I think, is that it encourages people to pay attention to details; hopefully it would encourage really detailed scenarios by proponents of Darwinism (ones that might be checked experimentally) and discourage just-so stories that leap over many steps without comment. So with those thoughts in mind, I offer the following tentative 'evolutionary' definition of irreducible complexity:

"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."

Anonymous said...

I wrote:
> And he never did demonstrate that any system was "IC" based on these newer criteria.

lee_merrill:
"That was most of his Black Box book, actually, demonstrating the implausibility of various steps that should be expected on all paths. Thus no need to presume a historical sequence, may it be noted."

That is *not* the case, lee_merrill. Behe's Black Box book only dealt with the first definition of IC: That of a system of interacting parts in which damage to any component destroys are severly damages the system's function. The later definitions added the concept that the IC system also must have arisen from several steps, none of which underwent selection. Behe was basically saying that an IC system which had to have arisen from a large number of selectively neutral transitions was less likely to evolve. Well, duh: That's basically *defining* an IC system as one that's unlikely to evolve. The problem with the additional criteria was that it now required Behe to demonstrate which transitions involved neutral steps and whether they were so improbable that they couldn't have happened in nature. That is something that Behe has *never* accomplished.

To summarize: Behe's original formulation of IC identified a subset of biological systems that met a *functional* criteria. It was his job to connect that to the relative evolvabilty of that class of systems. With the second verion of his IC definition he entangled a functional criterion with criteria related to historical pathways. Unfortunately, by considering evolvability in his newer criteria he's begging the question.

Torbjörn Larsson said...

Strange indeed, then, that the numbers seem to be coming from the IDers

It should be obvious that biologists presents such pathways not numbers. And that IDers don't as they always discuss a priori probabilities, probabilities which are besides the point as Unsympathetic Reader explains.