The Lenski et al 2003 paper, The evolutionary origin of complex features, is really worth reading. Here’s the abstract:
A long-standing challenge to evolutionary theory has been whether it can explain the origin of complex organismal features. We examined this issue using digital organisms—computer programs that self-replicate, mutate, compete and evolve. Populations of digital organisms often evolved the ability to perform complex logic functions requiring the coordinated execution of many genomic instructions. Complex functions evolved by building on simpler functions that had evolved earlier, provided that these were also selectively favoured. However, no particular intermediate stage was essential for evolving complex functions. The first genotypes able to perform complex functions differed from their non-performing parents by only one or two mutations, but differed from the ancestor by many mutations that were also crucial to the new functions. In some cases, mutations that were deleterious when they appeared served as stepping-stones in the evolution of complex features. These findings show how complex functions can originate by random mutation and natural selection.
The thing about a computer instantiation of evolution like AVIDA is that you can check back every lineage and examine the fitness of all precursors. Not only that, but you can choose your own environment, and how much selecting it does. There are some really key findings:
- Irreducibly complex functions evolve, where IC is defined as “if you take something away, it breaks”
- Functions evolve via Irreducibly Complex pathways, i.e. pathways in which there are several neutral, or even quite steeply deleterious steps.
- The more complex functions do not evolve without selection.
- All the functions evolve via at least some neutral steps
- There are many pathways to each function
- Not all the functions are achieved in the same way.
The AVIDA findings are very good evidence that there is no reason, in principle, why irreducibly complex things can’t evolve, that some degree of natural selection aka heritable variance in reproductive success in the current environment is necessary for for certain features to evolve, but that not all steps need be advantageous – because of the role of drift, even quite sharply deleterious steps can prove key to the emergence of a complex function.
Then there’s the Lenski lab e-coli experiments. But let’s start with AVIDA, because it really does rebut the idea that Irreducible Complexity, whether defined as the property of a function or of a path to a function, is an in-principle bar to evolution by natural selection.
They are not exclusive. In fact there is no possible evidence to disprove design. Every imaginable scenario is compatible with design.
The question is not whether design can be excluded; it is whether evolution is sufficient. AVIDA does not answer that, but it demonstrates that complex functionality can evolve.
Certainly none of the pathways to EQU are “direct”. They are all “indirect” by Behe’s definition, ie. the advantageous precursor steps are always advantageous for some function other than EQU. So co-option is part of what is happening.
The fact that that a precursor to an IC function could have a different function was always a problem for Behe’s original definition of IC, and Behe certainly concedes that “indirect, circuitous” pathways could lead to IC structures, although he considers that the more that are required the less probable the evolution of the IC structure.
But far more is going on here. Behe later refined his definition of IC (2000)
So Irreducible Complexity is now not an absolute property of a function, but a relative property of the necessary pathways to that function, i.e. the more unselected steps necessary to reach a function, the more IC it is.
And what AVIDA shows is that EQU readily evolves (e.g. in 23 out of 50 runs), always by pathways that involve a substantial number of steps that are not only unselected, but include actual deleterious steps. And while in theory only 16 mutations are necessary to produce the function, the pathways ranged from 51 to 721 steps. So while co-option occurs in evolution EQU (later functions use bits of earlier ones – examples are given in the paper), there are also a very large number steps in the pathways to each function that are not advantageous at all, and some that are quite sharply deleterious – in other words, regardless of what “degree IC” of the shortest necessary pathway to EQU is, the actual pathways by which it evolves are very high-degree IC.
The reason, of course, is drift.
Behe’s arguments (he does this in Edge of Evolution too) assume that necessary mutations have to be near-simultaneous to achieve the IC function, and that such co-incidences (i.e. literal coincidences) are hugely improbable, which indeed they would be if simultaneity was necessary.
But because of drift, it isn’t. Any mutation has a sporting chance of becoming quite highly prevalent in a population even if it is neutral or even slightly disadvantageous, and the more prevalent it becomes, the more chance it has of becoming more prevalent still. And that hugely increases the probability that a second or third of these necessary mutations will occur in the same individual.
AVIDA is a non-sexually reproducing system. In a sexually reproducing system, or one with HGT, the probabilities are higher still, because a mutation can propagate through the population independently of the genotype in which it first appeared.
Nonetheless, even without sex or HGT, drift ensures that EQU has a much higher chance of evolving than Behe’s principles would indicate, and not only that, but it does so via pathways that are far more deeply IC than the minimally IC pathway that exists.
And can evolve even if the function is IC, and by pathways that are high “degree IC” by Behe’s later definition
Not only does every step NOT need to be advantageous, pathways consisting of sequences of many non-advantageous steps and even deleterious steps regularly lead to the evolution of an IC function.
Interestingly, it shows yet another thing – that while you might think that an IC function would be very “fragile” (break one component and it doesn’t work), if it is highly advantageous it not only survives by “conservative” selection (i.e. breakages are eliminated) but continues to evolve – often re-acquiring functions its precursors lost over the course of the pathway to its first appearance.
Something Lenski also found in his e-coli studies. So while complexity is sometimes acquired by means of loss-of-function mutations, often those functions are later re-acquired (if they are useful). So AVIDA sinks yet another anti evolutionary argument! New functions are not always gained at the cost of loss of old ones – or, if they are – those old ones can be re-acquired, alongside the new one.
This is obviously true in the case of Avida.
Yes, by design.
Not sure if this is Serious-Mung or not. If it is, could you explain what you mean by both these comments?
From the paper:
Why could the ancestor not perform any logic functions? Well, that was of course, by design. By intention. By Godlike powers.
How does the program detect an EQU? I’m going to go out on a limb here and suggest it does so by design. Intentionally. By Godlike powers.
Oh, missed this comment from over the page.
No, the virtual organisms in AVIDA were not “designed” to perform logic functions (unless you count evolution as a designer).
The initial population (which could not perform logic functions) was designed.
So was the environment in which organisms that become able to perform logic functions got better access to resources.
But only the first generation were designed. All those that could perform logic functions were the result of evolution.
Often, in fact, in EAs, it is difficult to figure out how the evolved organisms are actually doing what they are doing. If they were designed to do it, the designer would know how. But the whole point of applied EAs is that you end up with a solution that you did not know in advance. The solutions are not the result of human design – they evolve, as in AVIDA.
Mung, as I said at the beginning: the original starting population, and the environment, are designed.
What is NOT designed, is the organisms themselves, beyond the first generation. All variants on the initial population are evolved, not designed.
If the ID argument is that a designer is needed for OoL and the provision of the pre-biotic earth, then they’ve got no beef with evolution.
But most ID arguments (and Behe’s specifically) is that some kinds of functions can’t evolve and have to be designed.
AVIDA shows that they can. Sure, they require a designed starting population and a designed environment. But Behe’s argument isn’t that complex functions can only evolve if you have a designer providing OoL and a prebiotic earth. He is arguing that even given those things, IC things can’t evolve.
It would appear that these potentialities were built in. By design.
Yes, the starting population and the environment were designed. As I said.
Hence the potentialities.
Evolution produced the actualities.
AVIDA does not rebut the argument that a Designer must have designed the starting population of life forms on earth, nor even life forms capable of translation and transcription of DNA into proteins.
What it rebuts is the argument that given the analog of those things, by Design or otherwise, IC functions can evolve and do so by way of deeply IC pathways, contrary to Behe’s case.
The mousetrap is created. So taking it apart makes it fail. Biology likewise.
However the computer example here is saying it created. yet all it did was work with already existing information. no creation at all.
Its hard to break it at all. since its just a equation adapting itself.
I think thats what fails it as a analogy for debunking IC.
Regarding IC, you say:
There is a difference between it having evolved and it having evolved by Darwinian means. The Mullerian two-step “add a component” sounds easy, but is it so easy?
In my opinion Darwinian evolution can tweek a feature to suit the environment as in Galapagos finches, but it is not capable of producing complex biological features.
Well I would say that, because we have no evidence of it existing, your less effecient motor is a “just so story”.
Darwinian evolution requires that it existed for the theory to hold up, therefore it must have existed. I brought up its efficiency because I think it is such a marvelous construction which, I think, should fill everyone with a sense of wonder.
CharlieM,
I’d say the same argument might be applied to your Designer, assembling the components from scratch (and sticking them on bacteria, including a whole range of diseases – thanks a bunch!).
Oh? Out of interest, how many types of such motors are you aware of?
One reason why AVIDA is a poor comparison of biological evolution is that with AVIDA co-option is simply a case of fitting the co-opted function into the existing instruction set. They can be directly spliced together and the product will either be advantageous or not. This is not the case with Behe’s IC example, the bacterial flagellum.
You cannot just take two separate components and join them together and expect they will be compatable (as my lawnmower example showed). And studying the flagellum shows this to good effect. For example the flagellar tail is driven through the hook but these two components cannot be attached directly. An adaptor ring is required to attach the one to the other. Even if these components are not an example of co-option it shows the difficulty of joining separate pieces to make a whole functioning system.
This joining is not a problem for AVIDA.
How many types of flagella motor are there Charlie?
OMagain,
Why does that matter? Pick any one and let’s look at it.
On discussing the resistance to chloroquine in malaria parasites Behe writes
(I think the above number should read 10 to the power 20 and not 1020)
So it is irrelevant whether the mutations are sequential or simultaneous because as Behe writes, he is looking at the actual data and not making calculations of probability.
I would say that AVIDA makes co-option not only possible nor even highly probable: it makes it inevitable.
Why?
It’s quite simple. Unless they are all exactly as efficient as each other then by definition some are less efficient then others.
IDers hate AVIDA because it reiterates that the evolution of irreducible complexity is not impossible in principle.
Since it’s not impossible in principle, whether it can evolve biologically is dependent on the specific biological fitness landscapes encountered in our world.
IDers are in the unenviable position of trying to characterize those fitness landscapes well enough to prove that the evolution of IC is impossible in biology, even though it isn’t impossible in principle, as AVIDA shows.
The interesting thing (well one of many interesting things) is that it even inverts the probability argument.
Behe thought that the greater the “degree-IC” the shortest necessary pathway to a function, the more improbable it would be for such a pathway to occur.
In the AVIDA runs, the shortest necessary pathway was never taken. Other, longer pathways occurred with greater frequency. i.e. high degree IC pathways are not less probable than lower ID degree pathways.
That’s really the problem for ID.
With Dembski they got all mathy and informationy. But when you follow the math, it doesn’t support ID.
Behe wrote an entire book devoted to the proposition that it is difficult to evolve a two step adaptation and statistically impossible to evolve two simultaneous two step adaptations.
And yet malaria did it in a human generation. Most likely by ignoring Behe’s route and finding another path.
The problem with ID is that all their best ideas are contradicted by inconvenient facts.
Why are there still IDists?
I guess that’s like asking “if evolution is true, why are there still monkeys”.
Well, we know why Behe is still an IDiot. Because that’s where he gets his money and fame.
No payoff or glamor for admitting he was wrong, no desire to settle for being just another mainstream biology professor.
The relative efficiency of extant flagella is not the point. It is assumed that these flagella have evolved from less efficient precursors. Where is the evidence for this?
Assuming, arguendo, that this happened, what kind of evidence would you expect to find, and why?
Do you have any references which confirm that IDers believe IC is impossible in principle?
What does the “I” stand for? What is the Edge in Behe’s Edge of evolution? Why has Behe asserted that some adaptations are unevolvable?
So it would appear that the digital organisms in Avida are designed to evolve the EQU logic function, that the potential to evolve the EQU logic function was inbuilt be the designers.
So yes, Avida shows us there an be multiple designers working together for the same purpose.
And if in-built potentialities being made actual is your idea of evolution, I’m all for it.
Welcome to Aristotelianism.
One
http://www.evolutionnews.org/2015/07/flagellar_diver097831.html
Sure. And the potential to solve the bacterial motility project was built into OoL and prebiotic earth.
That is irrelevant to the issue at stake here, which is Behe’s argument that IC functions cannot evolve, and not by deeply IC pathway (and that deeply IC functions are less probable than shorter ones).
If you want to make the argument that complex adaptations could only have evolved had a designer designed the simplest life-forms and prebiotic earth, that’s fine.
But in that case there’s nothing anti-Darwinian about your argument.
Charlie, the point of the AVIDA study is that the premise isn’t even necessary.
EQU, did not “evolve from less efficient precursors”. No precursor could perform EQU. There were precursors who could perform some other functions, and some of these were co-opted for EQU, destroying some of them in the process.
Indeed, some had to be destroyed BEFORE EQU could evolve, i.e. there were some actual deleterious precursors necessary.
But, then, because EQU is so useful (a bit like a flagellum), any mutation that destroyed EQU was quickly lost, and the survivors regained some of the functions they’d lost in order to acquire EQU.
Exactly the pattern of evolution that Lenski also found in his e-coli experiments.
In a similar way there is only one type of car, would you agree?
After all, all cars share a common set of core “car-proteins” right? Wheels, transmission, steering wheel.
Do you agree or disagree?
This paragraph of text existed before you wrote it and could have been found on Volume 32 on Shelf 5 of Wall 4 of Hexagon
Or here: https://libraryofbabel.info/bookmark.cgi?ycrprrvozpajsxyoxlgwfrl9
So, given that text existed before you wrote it, did you write it Mung or was the “potential” inbuilt by your designer?
If we evolved from monkeys, why are there still monkeys?
While I’m perfectly prepared to accept your answer, do you have a reference from an unbiased source?
The designer so loves killing children with dysentery and other diseases, that they are almost as diverse as beetles.
And the problem with some critics of Behe is that their criticism is contradicted by inconvenient facts.
Michael Behe:
CharlieM,
Behe gets the argument wrong, because Plasmodium is squeezed through a tiny bottleneck during its life cycle. The effective population size of Plasmodium is much closer to that of its local humans than it is to the census of all cells. This ‘edge of evolution’ can therefore be reached in quite small populations, irrespective of other ‘multiple-pathway’ and ‘Texas sharpshooter’ arguments.
We may find colonies of bacteria where they all had inefficient flagella because none of the colony members had stumbled on the way to make them more efficient.
But my point is that because there are no examples of simpler precursor flagella (the simplest known are still very complex and efficient by our standards), then we cannot just assume that they existed.
What does this have to do with my question? Have you read “The Edge of Evolution”?
Two things: first, malaria has moved on to have double drug resistance. This is the thing Behe said was statistically impossible. Double lotto winner. Behe is simply wrong. He’s wrong about AIDS also.
Regarding flagella, why would you expect to find inefficient versions in the current environment?
I have not read the Edge. I have read all of Behe’s replies to his critics. His argument is simple and it is wrong. He sees a path that appears to be blocked and concludes there is no usable path.
Can you tell me why the mutation rate in AVIDA is set so high? From my limited understanding I believe that the evolution of EQU is dependent on a fairly high mutation rate. What are the criteria for setting the mutation rate?
CharlieM,
We have no firmer grounds to say they didn’t. A central point of evolution is that it sweeps up after itself. History is written by the victors, etc etc. Fixation for one is extinction for another. Why should we always expect to find a precursor state, if it has been out-competed by another? The present absence of less-effective survivors is no guide to their historic non-existence. It is kind of expected.
Maybe a more serious question would be, Have apes and monkeys de-evolved from us?
From Sciencemag
First you conjure up some designer in your head, and then you blame your phantom designer for killing children.
Behe is the one who has said God designed malaria. You are the one who suggests that dysentery did not evolve without intervention.