Richard Dawkins is the Microsoft supported - Charles Simonyi Professor of the Public Understanding of Science at Oxford University and is perhaps the world's best known atheist.
Is he trustworthy in his presentation of the evidence that evolution can explain a complex structure like an eye?
In his book 1995 Book "River out of Eden" he says the following about a scientific paper written by Dan Nilsson and Susan Pelger:
[Their] task was to set up computer models ofevolving eyes to answer two questions ... [:] isthere a smooth gradient of change, from flat skin to full camera eye, such that every intermediate is an improvement? ... [and] how long would the necessary quantity of evolutionary change take?In their computer models, Nilsson and Pelger made no attempt to simulate the internal workings of cells.... Nilsson and Pelger began with a flat retina atop a flat pigment layer and surmounted by a flat, protective transparent layer. The transparent layer was allowed to undergo localized random mutations of its refractive index.They then let the model transform itself at random,constrained only by the requirement thatany change must be small and must be an improvement on what went before. The results were swift and decisive. A trajectoryof steadily mounting acuity led unhesitatingly from the flat beginning through a shallow indentation to a steadily deepening cup, as the shape of the model eye deformed itself on the computer screen... And then, almost like a conjuring trick, a portion of this transparent filling condensed into a local, spherical region of higher refractive index.... This ratio is called Mattiessen's ratio.Nilsson and Pelger's computer-simulation model homed in unerringly on Mattiessen's ratio.
As far as I know Dr Dawkins has made no attempt to retract or modify these sentences. They are seriously misleading. Dan Nilsson has stated in a letter sent to David Berlinski that there was no computer simulation in the research that led to the paper despite the fact that this simulation is brandished around the world as one of the most startling testimonies to the ability of random mutation and natural selection to produce complex structures like an eye!
This controversy blew up in the "Commentary" as "A Scientific Scandal" (available on the Discovery site abbreviated here)and David Berlinski has made available the letters and his response here.
If David Berlinski is to be believed the paper actually makes some fairly basic mathematical blunders.
Two things irritate me:
1. This serious case of misinformation has gone uncorrected for so long and is treated as a "minor error."
2. This kind of study fails to deal with the real issue which is in the area of a particular genetic program controlling the development of the eye and yet is presented as if the huge problems with making an eye have effectively been solved.
20 comments:
Well, I was on Dawkins' side til you pointed out he was sponsored by Microsoft... :P
Seriously though, if this guy has such a great rebuttal of an academic paper, why doesn't he send it in to a journal? I'm sure they'd love to hear from him. Unless he's worried about being caught out by the "crank filter" element of the scientific process?
It looks like he's got Dawkins bang to rights on using the wrong bit of technical phrasing but, as someone who's actually studying both mathematical models and stochastic models, I'd tend to interpret it as an innocent mistake. Apart from anything else, the only way they'll behave differently is that a stochastic evolutionary model can sometimes "tunnel" past obstacles to some extent. So using a mathematical model to represent a stochastic model is perfectly valid in terms of setting a lower bound on what the stochastic model is able to achieve. In short, the mistake Dawkins made actually makes his case slightly stronger, assuming that the mathematical model is accurate.
That last sentence is of course the key, and the model may indeed not be accurate - I'd need to check. I would, however note that the objections to Nilsson and Pelger's model were apparently raised after 1995, so Dawkins certainly can't be faulted for including it in a book published then. As such, calling him a "fraud" on that basis is extremely misleading - by that definition, every producer of science books in the world is a fraud.
Lifewish,
Dawkins made up a computer simulation where there was not one mentioned in the paper and as a result this idea that we have a computer simulation demonstrating the evolution of the eye is all over the internet.
Are you saying this is acceptable improvement of the public understanding of science?
Well, I was (independently) under the impression it was a simulation that I believe was done on a computer. So on that front I don't believe he made anything up. I thought that the area of dissent was whether it was a stochastic model or a mathematical one?
Lifewish,
As I understand it the paper reports a series of point calculations plotted as a graph. Is it seriously misleading to report this as if it were a well constructed computer simulation of biological change over time?
Not my field and, despite his intellect, I don't think these kind of programs are Dawkins field either.
As Lifewish has said, if he made an error it does of course need to be picked up and demonstrated as such. However, the use of Genetic Algorithms is a very well established area of study and the fact that Dawkins slipped up hardly throws the entire field into a quandary.
As I understand it the paper reports a series of point calculations plotted as a graph. Is it seriously misleading to report this as if it were a well constructed computer simulation of biological change over time?
Any computer simulation could be descriped as "a series of point calculations", and most could be plotted on a graph (or several, if my university projects are anything to go by :-/). As such, that is not enough in and of itself to convict.
I haven't actually read the paper (I'll attempt to look it up later today, if I can figure out where they're hiding the relevant journal) but from what I've read about it the only difference between their simulation and a true stochastic evolutionary model is that their version was designed to home in on the locally optimal and follow it, rather than taking the scattergun approach of evolution.
The advantage of this approach is that it takes a hell of a lot less processing power and behaves repeatably (so other biologists can check it out). The disadvantage is that it isn't necessarily perfectly representative of evolution but, like I said, this doesn't mean it would do better - quite the opposite, in fact. This evolutionary simulation (and yes, that appears to be what it was) couldn't do anything that evolution couldn't do given a decent population size*, and couldn't do a lot of stuff that evolution could do.
As such, I'd consider it valid to extrapolate from the statement that the computer model could achieve this evolution of the eye without assistance to the statement that natural organisms could achieve it. This conclusion is subject to change as and when I locate the damn paper.
Off-topic: I'm back at uni now, which means I'll (theoretically) be working a bit harder. As such, I'm probably going to be even more sporadic in my posting habits - sorry in advance.
* With sufficient population (and we're not talking ridiculous numbers here either) a stochastic model could be expected to explore pretty much the whole of the surrounding solution space. It would certainly locate the local best direction, and might even locate a better one that had to be "tunneled" through to. The only worry is whether it could achieve this as fast as this simulation, which can be compensated for appropriately.
According to the drawings in the Nilsson and Pelger paper, it appeared to me there were computer simulations involved.
Are you claiming there were none?
I won't take Berlinski's word for it, since he seems rather unable to stick to what the citations actually say (see the fantastic claims in his letter to Kansas newspapers last year).
And if there were no computer simulations, how does that invalidate the paper?
Which step of eye development they list in the paper do you claim is not actually observable in nature?
From what I understand non of the iterations of the Nilsson and Pelger's model reflect anything that ever appeared anywhere. They are just drawings on a piece of paper that reflect the out put of a mathematical model.
As a mathematics student myself, I take strong issue with your use of the word "just". The Nilsson and Pelger paper apparently demonstrates a plausible route by which complex eyes could evolve. No-one's saying it's the route by which the eye did evolve - the only implication of this paper is that even comparatively complex structures like bifocal eyes can evolve.
I'd note that this is more than sufficient to sink any "irreducible complexity" of the eye.
Lifewish,
I would disagree. The Nilsson Pelger paper does not deal with real biology. It does not explain how the new genetic information required to build an eye and use it originated at all. The eye requires complex developemental genetic pathways which are not touched on by this paper.
In a letter to David Berlinski the authors agree that there was no computer model but they were working on one and it was difficult. There is no mention of a computer model in the paper it was something that Dr. Dawkins added for the sake of his story.
It does not explain how the new genetic information required to build an eye and use it originated at all.
Uh... I'm having trouble reading this use of the word "information" without wincing. Won't someone please think of the mathmos?
I imagine that a lot of random mutations that varied the exact shape of the eye would have been floating around, none of them quite damaging enough to have been eliminated from the gene pool. It's fairly easy to change the structure of an eyeball - for example, I'm very short-sighted, which arises because my eyeballs are a different shape from normal. If I was a transitional species in the evolution of the eye, a difference in eyeball shape of that magnitude might give me significantly improved vision.
The eye requires complex developemental genetic pathways which are not touched on by this paper.
Having actually read the paper, you're quite right in saying that there was no actual computer model involved (except in the sense that they calculated the outcomes of their mathematical model on a computer). However, I'm not 100% sure what you mean by developmental pathways in this context as the paper quite clearly shows that there exists an evolutionary path to creation of the eye such that each step is an improvement.
This may or may not represent the actual development of the fish eyes in question (without access to a time machine or lots of fossil eyes - and eyes don't fossilise easily - there's no way of telling). However, the existence of such a pathway demonstrates that it is plausible for such an eye to evolve.
I post something that has several substantive comments and you focus on the word "just." Spoken like a true Darwinist. If you are not one yet, you are well on your way.
If I was a "Darwinist" in the derogatory sense in which the term is usually intended, I wouldn't be hanging out on this blog. I come here to challenge my views, not to affirm them - if anyone has a plausible disproof of evolution, I want to know about it first.
The reason I pointed to that particular word was because your primary argument appeared to rest on it. I can argue out the rest of your post if you'd like - I'm pretty sure that, for example, the final paragraph of your original post is completely inaccurate, for reasons that I shall address below.
Until Darwinists can point to actual data and not just fairy tale stories they should not be published in peer reviewed journals.
Eyes don't fossilise all that well, so we're never going to get a complete explanation as to how they formed. All we can do is try to figure out how they might have formed, and then demonstrate whether or not that possibility is plausible. Which, by remarkable coincidence, is precisely what Nilsson and Pelger have done in their paper.
Of course, if scientists were to be effectively banned from speculating about eye evolution for the reasons that you raise, ID proponents would presumably be effectively banned from speculating about the implausibility thereof. In which case the very existence of this thread is Bad And Wrong.
They know their drawing don't represent anything that ever existed.
Actually, it's specifically stated that the majority of "transitional forms" they describe in their article exist in one or more species today (see section 5 of the paper). I quote:
"Eyes closely resembling every part of the model sequence can be found among animals existing today (Salvini-Plawen & Mayr 1977). From comparative anatomy it is known that molluscs and annelids display a complete series of eye designs, from simple epidermal aggregations of photoreceptors to large and well-developed camera eyes (Salvini-Plawen & Mayr 1977; Land & Fernald 1992). The structural components of our model have counterparts with different embryological origin in different groups. For modelling purposes, these differences are irrelevant because selection for the various functions will operate on whatever tissue is present at the place where the function is needed."
So yes, the eyes in the drawings most definitely do exist - this is apparently not rocket science. The key point of the article is that these pre-existing forms can be legitimately seen as part of a coherent chain between basic light-detecting cells and complex eyes, with each step an improvement on the last.
Note that this completely undermines the final paragraph of your original post - you said:
"As with other essential body parts no eyes ever appeared anywhere except in the final fully operational form nor has there been anything new in 525 million years. Their model would predict thousands of transitional forms, some of which should be around today. Another dog barking in the night."
On reading the Nilsson & Pelger paper, it is obvious that the transitional forms you speak of do exist, if not by the thousand then at least by the dozen. However, since they themselves will be fairly locally optimal (not perfectly but close), they appear to be the finished product. This is standard with transitional forms - one way or another, a genuinely suboptimal species doesn't stay suboptimal for long.
People all over the internet as well as Dawkins and others use the Nilsson and Pelger article as a matter of faith and that it is a slam dunk that the eye has evolved. You should use your intellect and energies to help set the issue straight.
I hope that by actually looking up the article and critically analysing it I have indeed helped set the issue straight. I hope you too will look up the actual article and critically analyse it before derogating it further.
That is unless you worldview requires you to back a bankrupt theory such as Neo Darwinism.
My current worldview does indeed partially rely on the existence of the modern synthesis for validity. I would hope this means that, should "neodarwinism" be demonstrated to be bankrupt, I would in fact change my worldview rather than engaging in cognitive dissonance. Please pick me up on it if you feel I am not abiding by this.
Incidentally, it occurs to me that it must be rather hard for you guys to discuss this paper without having seen it. Of course, I can't send you a copy because that would be Evil Copyright Infringement and I'd go to jail for a gazillion years. As I said, I'm very keen to give you all the help in challenging my views, so this is a shame.
On a completely unrelated topic, do any of you have an email address you prefer to be sent random files at? If so, please get in touch with me at ajl59 at cam.ac.uk so I know where absolutely not to send a copy of any copyrighted material to.
I imagine that a lot of random mutations that varied the exact shape of the eye would have been floating around, none of them quite damaging enough to have been eliminated from the gene pool. It's fairly easy to change the structure of an eyeball -
The main problem underlying the evolutionist methods is exposed here. If there are random mutations that can change shape of the eye in a random manner then there must be also millions of random mutations that would deteriorate basic proteins resposible of making the basic structures forming the eye.
Why we only assume that mutation can hit the proteins that are responsible for changing shape of the eye?
Even the most basic life form has a DNA consisting 4,000,000 base pairs. In order to hit the correct sequence that represents the shape of eye the chance is 1/4,000,000. If somehow mutations can hit that specific code in one of the creatures then there is chance that it might hit irreveleant codes in other 3,999,999 creatures from same species. However the problem is even deeper. The sequence that must have been hit is also contained of special arrangement of amino acids. In order to slightly change the shape of the eye a few combination of aminoacids must change. How can we govern that the correct combination will change before any advantagous shape change can appear?
The chance to change only 10 of those basepairs would be 20e10. So we can roughly say that the chance of even the slightest advantegous random change in the eye without having any destructive effects would be:
20e10 x 4,000,000=4,096e19 or
1/40,960,000,000,000,000,000
It is the probabilistic fact that always overlooked by evolutionists. Don't forget it is the odd we are facing just at very begining of those many many tiny steps. Those astronomic odds must be overcomed at each of those so called advantegous steps.
In same way if we assume that nature can produce a unique and specific protein type for making any part of the eye say the Lens then the nature is also likely to produce random junk proteins (which %99 of them would be lethal) in random places of the eye structure.
But the difficulties are not limited to the above cases. Evolution is based on the selection mechanism i.e. Selection of the fittest. Assume nature can put a Lens in front of some light sensitive cells. Until the Lens can focus correctly and create a clear image what is the advantage of having a half-completed lens instead of a full one. From optics we know that focus can only happen in tight range of focal distances. Why we assume that all of these tiny steps (if there were any) are advantegeous and selectable by nature? Because we are tend to believe in it or it is an empirical fact proving this?
This problem is throughly discussed in Fred Hoyle's book of Mathematics of Evolution. He concluded that if there is such a thing called random mutations then the creature would cease to exist as a result of flood of bad mutations instead of evolving to any direction. Hoyle's book and calculations mostly ignored by evolutionists as normaly we would expect from them.
Moreover why don't we see the evolution in front of our eyes if the mutations are so frequent as darwinists used to believe? Why the basic structure of the eye has ramained intact for past billions of years among all creatures?
I just dont go in details of image processing part of the eye. Eye is not only a simple camera but the "seeing" is a process of decoding the color information and objects and interpreting them in an intelligent way. The most advanced image processing softwares available to us can't even simulate %0.001 of our vision capability. We can assume that Eye is evolved but our advanced image processing capability is also evolved?
Farshad,
Have you looked at Dembski's work on the probability of the production of the bacterial motor - the flagellum?
Yes I have looked at that calculations. Dembski end it up with something like 1/10e1170 as far as I remember for the flagellum.
Farshad,
Interestingly Ken Miller said that this calculation was wrong...Dembski then said...well do the calculation correctly then...I haven't found that Ken Miller has replied as yet.
Andrew,
Well, that's what Ken Miller is supposed to tell.
Probability calculations are never meant to be %100 correct. In probability we consider a model and base all of our calculations according to it.
If Dembski did a basic error in his calculations it would be so obvious that not only Miller but any college student could easily see that. By applying a different model Dembski could get a different result like 1/10e750 but as long as those results are so minuscule i.e. below 1/10e150 (universal probability bound) why should we be considered for accuracy of our model?
Astronomer Fred Hoyle hold a Phd. in mathematics and according to Darwinists he was also wrong. You can easily see in the wonderful world of neo-darwinism all mathematical models that refute the random mutation/selection fable are likely to be wrong.
It's not just a calculation error.
The premise of his calculations are flawed. Of course no one is going to do it "right" since the method is wrong. You can force a bad method to give you right answers.
"Information" is a vaguery in Demsbski's calculations. Even he admits, "This definition of information is highly abstract and by itself of little use to biology and science more generally."
http://www.utexas.edu/cola/depts/philosophy/faculty/koons/ntse/papers/Dembski.html
Read under "Complex information."
Does he give a rigorous definition of information? of complexity?
NO.
His definition is based on a poker hand the Royal Flush. Now please tell me why a Royal Flush constitutes information? Is this information more complex than any other straight flush?
"Although probabilities properly distinguish possibilities according to the information they contain, nonetheless probabilities remain an inconvenient way measuring information."
Then why, Dembski, do you persist in using probabilities to specify Complex Information(TM)?
Then in the very next sentence:
"We are clearly acquiring more information when we learn someone was dealt a royal flush than when we learn someone wasn't dealt a royal flush. And yet the probability of being dealt a royal flush (i.e., .000002) is minuscule compared to the probability of being dealt something other than a royal flush (i.e., .999998). Smaller probabilities signify more information, not less. "
WTF? Yes, clearly we have gained "information" when we LEARN that someone is dealt a Royal Flush, but is this the same information that is supposedly contained in the Royal Flush? This is intentionally confusing and conflated.
In one breath, so to speak, he says that probabilities are not a good way of measuring information, then says that small probablilties are good indicators of more information.
Do you see why this man is not taken seriously? Is this any kind of rigorous definition of Complex Information, which is a term Dembski coined?
Continuing:
"With a means of measuring information in hand..."
WTF? You just said...never mind...
"The distinction between specified and unspecified information may now be defined as follows: the actualization of a possibility (i.e., information) is specified if independently of the possibility's actualization, the possibility is identifiable via a pattern."
Ah, so a pattern is all you need to specify specified information eh?
But then subtly he attributes all patterns to " pattern-forming rational agents " , or does he? You might get that from reading the passage or you might not.
This is the nature of Dembski. He is never clear or about his definitions and when he goes about "explaining" his position, it is vague enough and unclear enough to shift around depending on who is asking questions or making criticisms.
Isn't it clear that he's being dishonest? Isn't it clear that he is a shyster, a scam artist?
"It is CSI that enables Maxwell's demon to outsmart a thermodynamic system tending towards thermal equilibrium." Sorry, Bill, even the Demon in constrained by the second law. Try again.
The url is:
http://www.utexas.edu/cola/depts
/philosophy/faculty/koons/
ntse/papers/Dembski.html
Thankyou beervolcano, you saved me a lot of typing.
There's an excellent summary of the simulation work in Mark Ridley's textbook "Evolution". It should be in any good library. In short, there was a simulation done.
Post a Comment