The Weasel Thread
by BilboI'm changing the title of this thread, and all of you can continue discussing whatever it is you want to discuss.
242 Responses to “The Weasel Thread”
Leave a Reply
You must be logged in to post a comment.
-
Featured Books
The Design Matrix: A Consilience of Clues by Mike Gene 
Your Inner Fish: A Journey into the 3.5-Billion-Year History of the Human Body 
Catalyzing Inquiry at the Interface of Computing and Biology 
System Modeling in Cellular Biology: From Concepts to Nuts and Bolts 
The Plausibility of Life By Marc W. Kirschner and John C. Gerhart 
Agents Under Fire by Angus Menuge 
Life's Solution by Simon Conway Morris 
Information Theory, Evolution and the Origin of Life by Hubert P. Yockey 
The Fifth Miracle by Paul Davies 
Nature, Design, and Science by Del Ratzsch 
Origination of Organismal Form by Muller & Newman 
Biased Embryos and Evolution by Wallace Arthur 
Rare Earth by Peter Ward and Donald Brownlee 
The Privileged Planet by Guillermo Gonzalez and Jay Richards 
The Way of the Cell by Franklin Harold 
The Volitional Brain by Benjamin Libet 
Evolution in Four Dimensions by Eva Jablonka & Marion Lamb 
The Evolution-Creation Struggle by Michael Ruse
September 15th, 2009 at 2:33 pm
Yeah, the irony is thick.
Don't hold your breath waiting for a disclaimer to appear, Bilbo. Behe has already made his choice clear: he is a Senior Fellow at the Discovery Institute Center for
the Renewal ofScience and Culture. He is squarely inside the big tent.Comment by olegt — September 15, 2009 @ 2:33 pm
September 15th, 2009 at 2:44 pm
He essentially did this with his first post .
Comment by Guts — September 15, 2009 @ 2:44 pm
September 15th, 2009 at 2:46 pm
Comment by Pez — September 15, 2009 @ 2:46 pm
September 15th, 2009 at 2:46 pm
I don't think Behe owes the world a disclaimer. He has stated his views on many occasions. Uniformity is the hobgoblin of critics.
Comment by Bradford — September 15, 2009 @ 2:46 pm
September 15th, 2009 at 2:46 pm
Uncommon Descent does in no way imply anti-common descent. It implies irregular or atypical common descent. CD via RM&NS is the typical neo-Darwinian held POV.
Comment by computerist — September 15, 2009 @ 2:46 pm
September 15th, 2009 at 2:59 pm
A disclaimer from Behe? Give me a break, you Americans are really a very funny people!
Comment by neddy — September 15, 2009 @ 2:59 pm
September 15th, 2009 at 3:41 pm
niwrad's latest is a gem. In software development error-checking and subsequent correction is a must. For internet protocols; UDP for example, used in socket programming for transmitter-reciever/client-server model is a connectionless oriented protocol, when applications such as streaming video or gaming or voice use UDP (a transport layer protocol), the data is simply sent out and without any re-transmitting on the other end primarily because it would take too much time to correct the faulty state (re-assembling the corrupt data due to things like EMI on the transmission medium) and an overlap on the current state would take place, subsequently scrambling the voice heard etc…, this making it a gamers, telephone callers or video watchers nightmare. UDP tolerates errors (commonly noticed by "lagging") but it is much faster and this makes it an excellent candidate for bandwidth intense app's that are real-time reliant. Some applications however, require errors to be checked and corrected, TCP over HTTP for example; data should be read correctly on TT's, so TCP makes sure that the data from the application layer maintains its integrity – retransmitting the data if necessary (for example; if checksum fails) until it is correct making it error-free and connection-oriented using basically back-and-forth acknowledgment and sequence numbers between the devices sending and receiving. For DNA coding I imagine data-integrity is a must, it is more of a TCP protocol then it is a UDP, the more I think about it the more this looks like a design choice to me, evolution via NS&RM would be better suited as a UDP protocol. In OOP languages such as C++ software developers use many methods for dealing with errors or unexpected events. Objects instantiated for example, can be checked for many possible error codes, such as "out of memory" or "invalid data" or "not enough video memory" to process/execute a file or function. Things can be done about the error, such as doing something else to solve the error before continuing or exiting the program. Some widely used techniques include simply if statements or exception handling using try catch blocks.
Comment by computerist — September 15, 2009 @ 3:41 pm
September 15th, 2009 at 3:47 pm
niwrad is simply staying with the latest creationist fad. In Paley's time, the idea of an advanced technology was a mechanical watch, nowadays it's computers. A cell of course is nothing like a watch or a computer.
Analogies aren't good for anything remotely resembling science. Make some predictions, guys, testable predictions. And no, "Darwinism won't explain this" doesn't count as a prediction.
Comment by olegt — September 15, 2009 @ 3:47 pm
September 15th, 2009 at 3:56 pm
If we knew nothing about how biology works, I bet a design hypothesis would make alot more accurate and testable predictions. Since Darwinism has hijacked design along the way then a design hypothesis is less likely to make "real" predictions, can't have it both ways it seems.
Comment by computerist — September 15, 2009 @ 3:56 pm
September 15th, 2009 at 3:57 pm
No kidding, it is much, much more sophisticated.
Comment by todd — September 15, 2009 @ 3:57 pm
September 15th, 2009 at 4:05 pm
Who died and left you god and king over the discussion of intra-cellular processes? As an engineer, I find it interesting. I'm not sure if you are aware of this, but the frontier of detailed understanding of intra-cellular process is relatively new field of study. Oh, and by the way, this is "Telic Thoughts". I put that in bold to help you remember. You and your ilk seem to be under the misguided, and yet mildly amusing, notion that everything telic-friendly people talk about is meant to impress the likes of people like you. Don't flatter yourself.
Comment by kornbelt888 — September 15, 2009 @ 4:05 pm
September 15th, 2009 at 4:16 pm
computerist, so it's the Darwinists's fault that IDers can't make any predictions?
Comment by olegt — September 15, 2009 @ 4:16 pm
September 15th, 2009 at 4:16 pm
Olegt:
Wow where do I begin, this is of course simply false. Analogy itself is used by researchers all the time. In fact, analogies are helping actual working biologists to understand various facets of cellular organization, specifically with regard to the TCP/IP protocol. You can read about it here . Network theory itself is currently flourishing because of the strong analogy between biological systems and human designs.
Dennis Bray is even suggesting that the cell is made of circuits that do computations just like electronic devices.
Comment by Guts — September 15, 2009 @ 4:16 pm
September 15th, 2009 at 4:25 pm
computerist Says:
September 15th, 2009 at 3:56 pm
What about your prediction about being able to show how to calculate FCSI? Have you given up on that claim?
Comment by Alan Fox — September 15, 2009 @ 4:25 pm
September 15th, 2009 at 4:25 pm
Re: Predictions:
here
Comment by Bradford — September 15, 2009 @ 4:25 pm
September 15th, 2009 at 4:30 pm
Buried Code
Interesting comments.
Comment by Bradford — September 15, 2009 @ 4:30 pm
September 15th, 2009 at 4:36 pm
Thanks for the link. From the review by Dave Rintoul
His insight that "it is much more difficult to infer internal structure from the observation of behavior than to create the structure that gives the behavior in the first place" is a powerful one, and should give pause to anyone who subscribes to the notion of "intelligent design", or who thinks that cellular activities are "irreducibly complex". Humans can be easily fooled into believing that human-like attributes can only be attributed to human-like intelligence.. But the notion that a cell is so complex that it must have been designed by a supernatural agent is similar to the response one might imagine if a caveman was confronted by a simple robot. In both cases the object seems beyond comprehension; in both cases the object can actually be described by simple physical laws, circuits and switches.
Comment by Alan Fox — September 15, 2009 @ 4:36 pm
September 15th, 2009 at 4:41 pm
Guts,
Systems biology, like string theory, is a young field that started out with a vague idea. While there is much activity in both of those new fields, it's too early to judge their success. Neither has really figured out how to make the vague ideas into a theoretical framework that would be useful in science. A recent nontechnical overview of the field—from a mainstream biologist's perspective—can be found here: B. Short, Cell biologists expand their networks, J. Cell. Bio. 186, 305 (2009). doi:10.1083/jcb.200907093
To the credit of systems biologists, there is quite a bit of data mining going on with various -omics. That is something that is sorely missing from the IDists' toolkit. No theory, no experiments, no data gathering.
Comment by olegt — September 15, 2009 @ 4:41 pm
September 15th, 2009 at 4:41 pm
Alan quoted:
That would be an excellent argument if IDers were insisting that in order to infer design, the design must be "incomprehensible" and impossible to design. They aren't.
Comment by Guts — September 15, 2009 @ 4:41 pm
September 15th, 2009 at 4:44 pm
Did you notice what was on Dave's reading list?
Comment by Alan Fox — September 15, 2009 @ 4:44 pm
September 15th, 2009 at 4:51 pm
Olegt:
Sure they have, check some of the books on the side panel. Many of the concepts can be applied to bacterial networks. With respect to the bow-tie analogy, it can even be observed . Barabasi in his book Linked said this regarding networks:
Comment by Guts — September 15, 2009 @ 4:51 pm
September 15th, 2009 at 4:53 pm
Come to think of it, didn't Dave review Mike Gene's book?
Comment by Alan Fox — September 15, 2009 @ 4:53 pm
September 15th, 2009 at 4:59 pm
Yeah Dave and Mike I think even debated awhile back.
Comment by Guts — September 15, 2009 @ 4:59 pm
September 15th, 2009 at 6:05 pm
Guts,
I know that Barabasi has been pushing this network analogy, but I am not all that impressed with the progress in that direction. The claims he makes in that passage are a bit, shall we say, grandiose:
Sorry, complexity is still a vague buzzword. Barabasi is an enthusiast of this direction, but we'll have to see how much progress in biology will be made along these lines.
Comment by olegt — September 15, 2009 @ 6:05 pm
September 15th, 2009 at 6:45 pm
Olegt:
We are already seeing it. It's not just Barabasi, work is being accomplished from all sorts of directions, Uri Alon comes to mind. Barabasi's work deals with preferential attachment. It starts with a small core of connected nodes, and then grows the networks by attaching new nodes to existing nodes with probability proportional to their indegree. In later versions of the model, some rewiring was also introduced to make sure that the graph that is generated is not acyclic. The basic model creates acyclic graphs because old nodes never link to new nodes.
So at first, he found that PA does not actually create networks with modularity. He then constructed a model that used a "copy and link" step which turned out to be extremely modular, or what he terms "hierarchical modularity", his model was used to study 43 organisms, it correctly predicted how the clustering coefficient (the average fraction of the neighbors of a node that are also neighbors of each other) followed a power law (see here ). Of course there's more work to be done, problems to be solved, etc, but thats the beauty of science.
Comment by Guts — September 15, 2009 @ 6:45 pm
September 15th, 2009 at 8:18 pm
As WEASEL so aptly proves.
Comment by Daniel Smith — September 15, 2009 @ 8:18 pm
September 15th, 2009 at 8:24 pm
Guts,
Have you seen this?
Can the genetic code be mathematically described?
Abstract:
Comment by Daniel Smith — September 15, 2009 @ 8:24 pm
September 15th, 2009 at 8:34 pm
Comment by Bradford — September 15, 2009 @ 8:34 pm
September 15th, 2009 at 8:50 pm
No, Dawkins's Weasel isn't science, it's a toy model used in a popular book about science. I just can't comprehend why
creationistsIDers invariably get excited every time it's dangled in front of them.Comment by olegt — September 15, 2009 @ 8:50 pm
September 15th, 2009 at 9:43 pm
Truthfully Olegt, Weasel is thought of as algorithmic proof by many for what is nearest and dearest to their hearts. In fact it was one of your own mainstream soulmates who first suggested that I blog on weasel about a month before computerist brought weasel front and center in a comment. This "nah it's just a toy" attitude would not have been adopted had your cohorts been successful at convincing others that weasel was a meaningful program. Reminds me of when I first started observing these discussions and I watched a PhD in chemistry posting comments in a group of mostly non-experts. He used to be fond of saying that enzymes did nothing more than accelerate the speed of reactions in discussions centered on the evolution of this or the origin of that. Finally I got fed up with this and pointed out that without a catalytic influence on biochemical reaction x the uncatalyzed version would consume more time than the life expectancy of the organism. He stoped the schtick but this showed me that anti-creationists, if you prefer that term, lack integrity and would rather show-up their despised opponents than convey a complete picture which accurately depicts what is going on.
Comment by Bradford — September 15, 2009 @ 9:43 pm
September 15th, 2009 at 9:45 pm
Bradford, for God's sake, reread what Dawkins said about Weasel and quit flogging it. I'm truly worried about your health. You spend too much time around the dead animal.
Comment by olegt — September 15, 2009 @ 9:45 pm
September 15th, 2009 at 9:57 pm
Bradford,
Here is your reading assignment for tonight:
http://en.wikipedia.org/wiki/W...
Pay attention to the last paragraph!
Comment by olegt — September 15, 2009 @ 9:57 pm
September 15th, 2009 at 10:09 pm
Olegt, if given the choice between shading the truth and convincing an audience that ID was bunk or being completely accurate and losing that opportunity my bet is that most swampies would opt for shading. It's easy enough to rationalize.
Comment by Bradford — September 15, 2009 @ 10:09 pm
September 15th, 2009 at 10:17 pm
From Olegt's link:
If I'm not mistaken it was Huxley- Darwin's Bulldog- who introduced the typing monkey imagery. Or is that incorrect?
Comment by Bradford — September 15, 2009 @ 10:17 pm
September 15th, 2009 at 10:30 pm
Wesley Elsberry:
http://www.antievolution.org/c...
Comment by Bradford — September 15, 2009 @ 10:30 pm
September 15th, 2009 at 10:43 pm
Bradford wrote:
Bradford, at least one of the swampies (myself) does not take position on whether ID is true or false. My take on it is succinctly summarized in Judge Jones's decision in the Dover case:
I can't add anything to it. ID's scenarios, like the tooth fairy, may be true but they are impossible to check. So they remain in the realm of invisible pink unicorns.
Comment by olegt — September 15, 2009 @ 10:43 pm
September 15th, 2009 at 11:58 pm
That summarizes my view of scientism and the belief that empirical testing will reveal anything meaningful about origins. Invisible pink unicorns with a good deal of hubris and self-delusion thrown into the mix.
Comment by Bradford — September 15, 2009 @ 11:58 pm
September 16th, 2009 at 12:04 am
Actually my question for Olegt was not about the truth or falseness of ID but rather to what extent its opponents will go toward trying to discredit the notion. Would they withhold information, of which they have knowledge, in order to advance the case against what is to them an odious idea?
Comment by Bradford — September 16, 2009 @ 12:04 am
September 16th, 2009 at 3:03 am
That is rather an unwarranted smear. The only really odious aspect about ID was the claim that there was substantive material entitling the subject to be taught as science in public schools.
Even arch-villain Richard Dawkins says if there were genuine evidence for God, he would have to accept it. What is the point in sticking with a theory that is demonstrably false. No scientist can do that successfully, as anyone can examine and present evidence. The Catholic Church tried to keep Galileo under control. Unfortunately, they couldn't hide the movement of planetary bodies from view.
Today, you have the internet. You have ISCID. If anyone has a good ID theory they can publish. If anyone has a good ID theory, you could publish it!
Comment by Alan Fox — September 16, 2009 @ 3:03 am
September 16th, 2009 at 10:31 am
Daniel:
Mike Gene wrote about this in his book, one of the many reasons I would give the genetic code a very high Analogy score.
Comment by Guts — September 16, 2009 @ 10:31 am
September 16th, 2009 at 10:54 am
“niwrad's latest is a gem…”
Interesting analogy. The TCP application layer ensures compatibility by controlling the formatting and transfer of data per end-user specs. Who or what is the “end-user” in this analogy?
(Are disclaimers obligatory now? Do I have to issue a disclaimer for asking a question?)
Comment by Rock — September 16, 2009 @ 10:54 am
September 16th, 2009 at 10:55 am
And what about "compatibility"?
Comment by Rock — September 16, 2009 @ 10:55 am
September 16th, 2009 at 11:49 am
My analogy was to show that instead of the expected evolution that "admits" errors passively (more like a UDP protocol) we have an evolution that controls and prevents errors (more like a TCP protocol). That kind of beats the whole point of Darwinian Evolution. Natural selection itself, doing what it does best (according to Darwinists) would more likely select mechanisms that promote random errors and leave them subject to brute-force. This is truly unexpected from a theory as well tested as gravity.
Comment by computerist — September 16, 2009 @ 11:49 am
September 16th, 2009 at 12:43 pm
Moreover Darwinian Evolution seems to posit that initial stages guided by a UDP like protocol evolve (inevitably?) toward systems which function like a TCP protocol.
Comment by Bradford — September 16, 2009 @ 12:43 pm
September 16th, 2009 at 1:27 pm
Are my comments invisible to you computerist? Just put me out of my misery. Have you abandoned your claim to be able to calculate FCSI? Confession is good for the soul, remember!
Comment by Alan Fox — September 16, 2009 @ 1:27 pm
September 16th, 2009 at 1:43 pm
Alan Fox, has it occurred to you that analyzing the effects of error controls can be part of an FCSI developmental process? What is it I am continually told when I ask questions?- something like this is an active field of investigation.
Comment by Bradford — September 16, 2009 @ 1:43 pm
September 16th, 2009 at 2:53 pm
Unless of course nothing less than METHINKS IT IS LIKE A WEASEL is selectable as "fit". Then monkeys on typewriters and cumulative selection are equal. (And this is why the analogy to anything even remotely resembling real evolution fails).
WEASEL even fails as a toy analogy.
Comment by Daniel Smith — September 16, 2009 @ 2:53 pm
September 16th, 2009 at 3:07 pm
No, Bradford. But a process whereby FCSI gets developed into something meaningful sounds like progress.
What say you, computerist?
Is this what you mean by Error control?
Divided as we are by a common language, I have to say I am baffled as to why you are saying this to me. Could you explain or is there an earlier comment I have missed you can link to?
Comment by Alan Fox — September 16, 2009 @ 3:07 pm
September 16th, 2009 at 3:16 pm
Dan asserts:
That may be your assessment, Dan, but I find it improves with age. Zachriel and others' updating of Weasel have recently made the point that Dawkins was trying to illustrate about cumulative selection as opposed to "tornado in a junkyard" much clearer. I should also commend Rev. Gordon Elliot Mullings for his sterling rôle as devil's advocate with his comments at Uncommon Descent.
Comment by Alan Fox — September 16, 2009 @ 3:16 pm
September 16th, 2009 at 3:21 pm
Unless readers scroll down to his first post they would miss it. What's wrong with including it in his profile?
I fail to see the joke.
And thanks everybody else for staying on topic.
Comment by Bilbo — September 16, 2009 @ 3:21 pm
September 16th, 2009 at 6:04 pm
Truthfully, Bradford, what weasel does is algorithmic disproof of what IDists hold nearest and dearest to their hearts.
I suggested you blog weasel because you brought it up and obviously misunderstood it. Naively, I thought an explicit discussion would help you understand it better, but obviously I was mistaken.
For example, it is just a toy, but you continue to think someone thinks it's more than that. It's a simple toy that clearly and concisely proves ID's "random processes can't create information" claim false.
Comment by don provan — September 16, 2009 @ 6:04 pm
September 16th, 2009 at 8:32 pm
There you have it!
Given a fixed target, lots of direction and no leeway in selection criteria, random processes can create information.
What were you IDers thinking?
Comment by Daniel Smith — September 16, 2009 @ 8:32 pm
September 16th, 2009 at 10:25 pm
Dumb.
Comment by Bradford — September 16, 2009 @ 10:25 pm
September 16th, 2009 at 10:46 pm
Wow, the poor weasel is still being kicked around this blog? You, guys, are a riot.
Comment by olegt — September 16, 2009 @ 10:46 pm
September 17th, 2009 at 2:20 am
And when they finally beat Weasel to death, they will turn to biomorphs. The original code is still available, something for Gordon Mullings to get his teeth into.
Comment by Alan Fox — September 17, 2009 @ 2:20 am
September 17th, 2009 at 3:02 am
You're so clever, Alan, knowing his real name and all. Way to go.
Comment by Pez — September 17, 2009 @ 3:02 am
September 17th, 2009 at 3:41 am
Not at all, Pez, though thanks for the compliment.
All any onlooker has to do is follow "the always linked" where Rev. Mullings publishes his name quite openly.
Comment by Alan Fox — September 17, 2009 @ 3:41 am
September 17th, 2009 at 6:45 am
Yeah, keep it up and bat your eyes. Very impressive. Boy, you sure know a thing or two.
Comment by Pez — September 17, 2009 @ 6:45 am
September 17th, 2009 at 9:04 am
Then the presence of FCSI is not a conclusive indication that a telic source was involved. Glad we have that sorted out.
A target is needed for string fitness because string fitness is established only by human convention. There is no physical law that says WEASEL is well-formed but WQASEL is not. Suppose we remove the target, as I proposed here. We're still brewing in FCSI via random processes, aren't we?
Comment by Tom MH — September 17, 2009 @ 9:04 am
September 17th, 2009 at 9:11 am
The lesson is that if targets are preset to ensure they are hit then intelligent design is evident.
Comment by Bradford — September 17, 2009 @ 9:11 am
September 17th, 2009 at 9:28 am
olegt:
Hello olegt, I've been trying my best to follow along on the subject of "weasel" just to see what point it was originally meant to convey. Perhaps, being from the "other side," "far side," or "dark side," you could concisely explain to me what message "weasel" is supposed to convey. What proof of concept is weasel meant to provide?
Comment by CJYman — September 17, 2009 @ 9:28 am
September 17th, 2009 at 9:29 am
And if there is no preset target (per the *ahem* second part of my post)?
Comment by Tom MH — September 17, 2009 @ 9:29 am
September 17th, 2009 at 9:42 am
It’s turned into a Creationists’ Circle Jerk! That weasel has been flogged so raw as to be unrecognizable.
Have you ever felt like you were losing control, Bilbo? Or can you pinpoint where you began to lose control?
Scientists have to be able to ask questions w/o issuing idiotic obligatory “disclaimers.”
They are shibboleths, chilling in their efffect, and anti-scientific–cutting off free expression and free exchange of information, ideas, beliefs, w/o which it is impossible to do science.
Comment by Rock — September 17, 2009 @ 9:42 am
September 17th, 2009 at 9:49 am
Presetting a target to ensure it is reached is a blatent example of design. Nudging is more subtle.
Comment by Bradford — September 17, 2009 @ 9:49 am
September 17th, 2009 at 10:02 am
CJYman wrote:
CJYman, I have explained what Weasel does and does not many, many times on this board, but is seems to fall on deaf ears (Bradford is a great example). I no longer care to do so. Read Dawkins's book, he explains it very clearly. Or if you don't care to get the book, the Wikipedia article has all the salient points.
Comment by olegt — September 17, 2009 @ 10:02 am
September 17th, 2009 at 10:06 am
CJYman, I read what is said about weasel. It is an illustration of the power of design to effect a predetermined outcome.
Comment by Bradford — September 17, 2009 @ 10:06 am
September 17th, 2009 at 10:14 am
Bradford wrote:
CJYman, go ahead, read the original description by Dawkins, and then reread Bradford's comment. I hope you will see my point.
Comment by olegt — September 17, 2009 @ 10:14 am
September 17th, 2009 at 10:25 am
The description by Dawkins is not the telling part of this. It is weasel itself. Weasel shows the power of a designed process having a preconceived target. No disclaimers necessary.
Comment by Bradford — September 17, 2009 @ 10:25 am
September 17th, 2009 at 10:42 am
The way I am reading this discussion about Weasel is that Bradford and his ID friends simply cannot accept, for psychological or ideological reasons I cannot or dare not fathom, that Weasel is a good demonstration of the relative speed of a cumulative selection-based search compared to a random search.
Comment by Hammerstein — September 17, 2009 @ 10:42 am
September 17th, 2009 at 10:51 am
Hammerstein, it is pointless to resist.
Comment by olegt — September 17, 2009 @ 10:51 am
September 17th, 2009 at 10:51 am
Hammerstein you gotta do some work to take in what was written. Early on I pointed out that the specified outcome was front loaded into the programming. That's an inherently teleological approach and its biological counterpart has been frequently advanced at this blog.
The advantage of a selection based search is not disputed.
Comment by Bradford — September 17, 2009 @ 10:51 am
September 17th, 2009 at 11:36 am
Begging the question.
Comment by Guts — September 17, 2009 @ 11:36 am
September 17th, 2009 at 12:31 pm
Guts wrote:
Nope. Just describing the standard model of evolution.
Comment by olegt — September 17, 2009 @ 12:31 pm
September 17th, 2009 at 12:33 pm
Nope, he's making an assertion about reality, which is questionable.
Comment by Guts — September 17, 2009 @ 12:33 pm
September 17th, 2009 at 12:48 pm
Guts wrote:
Welcome to the world of science, Guts. Scientists do make assertions about reality. They are called theories.
Take the theory of Brownian motion. It models collisions of the Brownian particle with molecules as a random force of a certain strength (white noise with Gaussian correlations, to be exact). You may question the assumption of randomness all you want, but the theory makes a nontrivial prediction that is born out experimentally, so the standard model of Brownian motion is generally accepted.
Comment by olegt — September 17, 2009 @ 12:48 pm
September 17th, 2009 at 12:51 pm
Bradford, you seem to be ignoring me.
Well?
Comment by Tom MH — September 17, 2009 @ 12:51 pm
September 17th, 2009 at 12:54 pm
Do you dispute that a selection-based search has an advantage even if there is no target?
Comment by Tom MH — September 17, 2009 @ 12:54 pm
September 17th, 2009 at 12:57 pm
Olegt:
Indeed, and it has lead the "standard model" to become "surprised" by such discoveries as this , which are more and more frequently turning up. Incidently, evolutionary theory is itself changing , this will continue to happen such that it would be, at least, more difficult to conjure up the will to make Dawkins's assertion in public, I predict.
Comment by Guts — September 17, 2009 @ 12:57 pm
September 17th, 2009 at 1:20 pm
Tom, I thought you would gather that my nudging link to Mike Gene's blog entry was an indiicator that a teleological process could be feasible without a preset target although a directional indicator is embedded in front loading. Think of it as the difference between moving in the direction of a target and a guarantee that you will hit the target.
Comment by Bradford — September 17, 2009 @ 1:20 pm
September 17th, 2009 at 2:03 pm
Bradford, thanks for the link. I have no trouble with the idea of the feasibility of teleological processes, I just don't see how they are necessary. But I'm intrigued, and I promise to give Mike Gene's post (and some of the linked soucrcs) a closer read.
What I was getting at, however, is the ability of RV+NS to efficiently find a fitness peak even if the configuration for that fitness peak (the "target") is unknown at the outset. That would suggest that FCSI can indeed be generated by non-telic processes. What part of this (if any) do you reject? (Note that one's persepctive on front-loading is not at issue.)
Finding fitness peaks is what evolutionary computation is all about. We don't use it in engineering to replicate designs we already knew in advance — why would we? The design solutions are novel. Where do you suppose the FCSI in them came from?
Comment by Tom MH — September 17, 2009 @ 2:03 pm
September 17th, 2009 at 2:39 pm
Yes it is, by the neo-Darwinists who insist that cumulative selection produces FCSI.
Biological processes and functions simply outweigh the clues nature can provide making cumulative selection obsolete for the task with exception too acting on pre-existing FCSI (ie: varying function; take your pick for a weasel without a target on whether it increases or decreases function).
Comment by computerist — September 17, 2009 @ 2:39 pm
September 17th, 2009 at 5:11 pm
computerist wrote:
I don't think it does: FCSI is not a well defined quantity. No one has ever been able to explain how it can be measured experimentally.
Comment by olegt — September 17, 2009 @ 5:11 pm
September 17th, 2009 at 5:33 pm
Weasel without a target would be like evolution without survival.
Comment by don provan — September 17, 2009 @ 5:33 pm
September 17th, 2009 at 5:37 pm
I could justifiably send most of the comments here to the memory hole, but I think I'll do something else instead.
Comment by Bilbo — September 17, 2009 @ 5:37 pm
September 17th, 2009 at 8:40 pm
Tom, we need to make clear distinctions between telic and non-telic processes (probably a good blog topic). There is also the matter of function about which Mike has made some insightful comments at his blog. You write that one's perspective on front loading is not an issue but I think it is the central issue where telic processes are concerned. In principle I will not rule out a non-telic process found to generate FCSI but since I believe the values of the basic forces of nature were front loaded at the inception of the universe, in a very real sense all processes are telic AFAIC. So I would retreat to a comment by LaPlace to credit any process as non-telic. That I have blogged about before.
Comment by Bradford — September 17, 2009 @ 8:40 pm
September 18th, 2009 at 9:47 am
Talk about hijacking a thread! Sorry, Bilbo…
Comment by Tom MH — September 18, 2009 @ 9:47 am
September 18th, 2009 at 10:37 am
A position that seems consistent with theistic evolution…but now we have left the realm of science (unless you can derive some testable predictions from this hypothesis*) and entered the realm of philosophy.
[*I am not saying you can't, I just don't see it.]
There is also a significant amount of air between the above position and the position that claims that RV+NS are insufficient to account for the machinery of life. IOW the Master Billiards Player was unable to call all N pocket+ball pairings prior to the break, and has to reach in now and then to steer the balls into their designated pockets as the shot progresses. The "front-loading" of establishing the universal physical constants (your position) was insufficient to guide the emergence of lide in the universe.
[The "RV+NS is not enough" view seems to be tied up with religious viewpoints, including a not-too-subtle desire to secure scientific acknowledgment of the statements in the Book Of Genesis as literal truth. Personally, I have never understood why a deist would argue that God could not predict -- and even intend -- the outcome of a dynamic N-body problem in four forces with random events throughout. Seems right up the ally for an omnipotent being. The meddlesome God reaching in to force each ball into its designated pocket is not a very inspiring billiards player. JMHO.]
The lesson of weasel (it feels so good to finally be on topic!) does not address, or even attempt to address, your philosophical viewpoint of the origin of the universe and the physical properities that made life possible, It does not even address the probability of the emergence of life. It only treats, by way of analogy, the subject of the ability of RV+NS to rapidly acquire peaks of fitness, and if you extend the analogy (as I have done) to something like protein catalytic rates then you can understand how highly catalytic proteins could evolve and not need to be assembled tornado-in-a-junkyard fashion by random chance. THAT'S IT.
The use of a target in weasel is irrelevant; it is required only because the analogy weasel uses is string "fitness" (correspondence to a known Englidh phrase) and string fitness is defined by human convention. The search for protein fitness (catalytic capability) does not require a target. Protein fitness is defined by the properties of organic chemicals, which may indeed be (as you contend) the evidence of telic forethought, but that is entirely outside the scope of the subject weasel is addressing.
Fair enough. As a newcomer to TT, I don't wish to run the rest of you through what may already be well-covered ground.
And a nod again to Bilbo for being gracious enough to let us kick the weasel around on his thread.
Comment by Tom MH — September 18, 2009 @ 10:37 am
September 18th, 2009 at 11:05 am
I wonder why?
Here's the typical argument:
Weasel is nothing. It makes no claims to anything even remotely resembling real evolution. It's just a toy. What's the big deal? All it does is show the power of one of the core mechanisms of the theory of evolution: cumulative selection, and prove one of the central tenets of ID wrong: that random processes cannot generate complex specified information. That's it! See it's really nothing – less than nothing. Why does it upset you so?
Comment by Daniel Smith — September 18, 2009 @ 11:05 am
September 18th, 2009 at 1:27 pm
Sorry. I made it starting here and in the following post. An argument by analogy, and therefore (admittedly) a not very good one.
Yup. Finding ways to avoid the low hill trap is part of evolutionary computation. One never knows if the current optimum is the best, so the search process is often re-seeded randomly across the state space, rather than just a making small variations to the current best solution.
BTW this is a design problem as well. We never know if the prevailing engineering solutions are optimum in a global sense. Some business entities put much effort into creative ideation ("brainstorming" and so forth) to try and break the current design paradigm and find better alternatives. If they exist. It's a high risk, high payoff process…and intensely maddening to try to explain to supervisors who want to see detailed progress agianst detailed milestone schedules, earned value reports, etc. etc.
Your expertese here undoubtedly exceeds my own, which isn't much (I am not a biochemist). If the fitness landscape is "Cantorian" then RV+NS will work no better than random draw. But if there is a narrow winding pathway to the top, evolutionary computation can be devilishly efficient in finding it.
Comment by Tom MH — September 18, 2009 @ 1:27 pm
September 18th, 2009 at 1:46 pm
Thanks
Comment by Guts — September 18, 2009 @ 1:46 pm
September 18th, 2009 at 5:26 pm
Not if the domain is finite, which is usually the case in biology.
Comment by chunkdz — September 18, 2009 @ 5:26 pm
September 18th, 2009 at 7:29 pm
This, to me, epitomizes the differences between real-world evolution and evolutionary algorithms used in design. Real-world evolution is limited by so many factors that a realistic real-world evolution algorithm would probably be useless for design. Design algorithms work much better when they can essentially explore all options.
Just curious Tom, how much time have you invested in the study of actual cases of biological evolution? I'm talking about natural selection (in the lab or in the real world), not artificial selection.
Comment by Daniel Smith — September 18, 2009 @ 7:29 pm
September 19th, 2009 at 2:33 am
The Weasel program shows cumulative selection. (But so does stacking blue blocks one on top of another to reach a predetermined height, as any little kid knows. When you choose something you select it. And you can keep doing this progressively.) Other than that it has nothing to do with real world biological evolution.
Hasn't the weasel been beaten to death?
Comment by kornbelt888 — September 19, 2009 @ 2:33 am
September 19th, 2009 at 6:32 am
Kornbelt888:
I would ahve thought so, but Demski and Marks recent paper misunderstanding how the program works seems to have produced the recent flurry of comment.
CJYman:
Here is an explanation by Patrick May which might help.
It isn't a proof of anything. Dawkins came up with it as an analogy to demonstrate the power of cumulative selection in his 1986 book "The Blind Watchmaker" It is odd that his biomorph programs haven't achieved the same notoriety.
Comment by Alan Fox — September 19, 2009 @ 6:32 am
September 19th, 2009 at 6:44 am
PS to Bilbo,
On glancing through the thread, it seems I might have contributed a few off-topic comments. So, sorry!
But, anyway, computerists started it!!!
Comment by Alan Fox — September 19, 2009 @ 6:44 am
September 19th, 2009 at 8:45 am
I'm not sure I catch the sense of "Not" in your comment. Too many negatives in the antecedants!
Engineering and biology are constrained in different ways, but their domains are rather poorly delimited (and may be crossing: bio-engineering). In any case, apologies for not understanding you.
Comment by Tom MH — September 19, 2009 @ 8:45 am
September 19th, 2009 at 8:56 am
Trust me, engineers are never free to explore all the options – we have a budget and a schedule to meet!
Many of the pioneers in evolutionary computation are ruthless in trying to eliminate artifacts of human intelligence in their decision-making algorithms. "Expert Systems" is an epithet. Real-world evolution is considered a gold standard.
None. I have a science education but I am a career engineer.
Comment by Tom MH — September 19, 2009 @ 8:56 am
September 19th, 2009 at 11:56 am
Ah but they obviously have a skewed view of what "real-world evolution" actually entails. They are much more likely to get results with computations that are much looser in their fitness functions (allowing more options to be tested) than real-world organisms can withstand.
That explains why you would think highly efficient enzymes could be easily evolved.
Comment by Daniel Smith — September 19, 2009 @ 11:56 am
September 19th, 2009 at 2:33 pm
Many of the pioneers in evolutionary computation are ruthless in trying to eliminate artifacts of human intelligence in their decision-making algorithms. "Expert Systems" is an epithet. Real-world evolution is considered a gold standard. Comment by Tom MH — September 19, 2009 @ 8:56 am
Actually the design of expert systems has little to do with evolutionary computing, and it is quite literally impossible for us to eliminate the idiosyncracies of human thought processes from anything we do. (Reduce? Possible. Eliminate? Not.) As is usually the case, we do not rely on computer programs to make critical decisions, particularly about what problems (the input) are to be solved by the program and whatever solutions it outputs are put in the hands of those very-human experts to evaluate and act upon accordingly. We are all “control freaks” in that way. Computers are never allowed to make such fateful decisions, as in the case of medical diagnosis, e.g., where expert systems have had such success.
Also, “real-world evolution” (whatever that process may be) is not any kind of “gold-standard” of evolutionary computing, but rather its inspiration in the most basic terms. Standards in evolutionary computing are standards devised by computer scientists (not biologists), such as job-shop scheduling and traveling salesman (and many more much more arcane standards). Engineers are no doubt impressed and inspired by the (sometimes) brilliant, innovative, and amazingly effective designs produced by evolution. They do not attempt to systematically reproduce such results, and mostly evolutionary computing is about self-adapting programs, which puts it at odds with both IDers and Neo-Darwinists. IDers believe life forms are adapted (designed) by God, and Neo-Darwinists believe they are adapted at random.
There is a group of evolutionary “computists” who do attempt to recreate naturally occurring processes of biological evolution, and they are population genetics theorists. It is increasingly more common for them to test their own (and others’) theories about biological evolution via computer simulations, evolutionary computing.
Some effort has been invested in these programs to make them “realistic” and if one seeks sourcecode for programs then H. Allen Orr, or Sean Rice, or John Gillespie is certainly the preferred choice to Richard Dawkins.
I mean anyone other than IDers would prefer just about anyone other than Richard Dawkins (a favorite whipping-boy around here). Always choose the easiest target to hit. Right, Bradford?
Comment by Rock — September 19, 2009 @ 2:33 pm
September 19th, 2009 at 2:43 pm
The description by Dawkins is not the telling part of this. It is weasel itself. Weasel shows the power of a designed process having a preconceived target. No disclaimers necessary.
Comment by Bradford — September 17, 2009 @ 10:25 am
Weasel is a lousy program all the way round, Bradford. It neither demonstrates the power of selection nor the power of design.
Comment by Rock — September 19, 2009 @ 2:43 pm
September 19th, 2009 at 2:57 pm
How about the power of poor design?
Comment by Daniel Smith — September 19, 2009 @ 2:57 pm
September 20th, 2009 at 10:04 am
The Original WEASEL(s)
Comment by Bradford — September 20, 2009 @ 10:04 am
September 20th, 2009 at 10:06 am
Simply Not Credible
Comment by Bradford — September 20, 2009 @ 10:06 am
September 20th, 2009 at 10:48 am
(From Bradford's link to UD)
Bill Demski writes:
Has Dembski ever troubled to contact Dawkins and ask him if he happened to keep a copy of his original Apple Basic code written at least 23 years ago? I predict not, but perhaps if we all send nice emails to Richard Dawkins he might think it important enough to check his attic and see if he still has the tape cassette or whatever it was programs were stored on a quarter century ago.
And, at the end of the day, Dembski is attempting to poke holes in a model that nobody is claiming is demonstrating anything more than cumulative selection as opposed to all-in-one-go. Demonstrating flaws in evolutionary theory does not give a free pass to ID, so how much less point is there in attacking a model that no one considers to be any sort of support for evolutionary theory. Maybe it is because it is fairly clear to most fair-minded observers that Dembski misunderstood Weasel, so is now trying to pull another Weasel out of the hat to fit his misunderstanding. Doesn't do much for his remaining integrity.
Comment by Alan Fox — September 20, 2009 @ 10:48 am
September 20th, 2009 at 11:11 am
What is the evidence that Dembski did not understand Weasel? If it is not strong evidence then it is not Dembski's integrity which should be questioned. The integrity of his critics is the issue.
Comment by Bradford — September 20, 2009 @ 11:11 am
September 20th, 2009 at 11:25 am
The fact that he and Marks claim Weasel is a partitioned when it plainly wasn't and the fact they claim correct letters "fix" or"latch" when they plainly don't. Many people have been able to reconstruct their own weasel versions from Dawkins' description in "Blind Watchmaker", but Dembski has turned a deaf ear and now appears to champion one that plainly does not tally with Dawkins' original.
What other motive would you suggest for looking for a program that fits with his and Marks paper, rather than the correct one which I suspect he is hoping will not surface. Let's see if Richard Dawkins can find anything in his attic.
Comment by Alan Fox — September 20, 2009 @ 11:25 am
September 20th, 2009 at 11:26 am
Wesley Elsberry has been pointing out Dembski's error out to him from around 1999 and been utterly ignored by Dembski. Who lacks integrity?
Comment by Alan Fox — September 20, 2009 @ 11:26 am
September 20th, 2009 at 11:27 am
The letters are fixed by their suitability to the target. That's why this demonstrates a designed objective rather than a selection process. What is selected correlates to objectives conceived at the outset. That's design pure and simple.
Comment by Bradford — September 20, 2009 @ 11:27 am
September 20th, 2009 at 11:35 am
Is this an accurate depiction of Weasel or not?
Comment by Bradford — September 20, 2009 @ 11:35 am
September 20th, 2009 at 11:37 am
Oh give me strength, Bradford!
There is a specific issue here. Weasel does not work in the Dembski and Marks claim it does. Rather than admit that they have misunderstood Weasel, they, or Dembski at least, are prepared to take an obvious fake contribution to his website and claim victory.
Comment by Alan Fox — September 20, 2009 @ 11:37 am
September 20th, 2009 at 11:43 am
I think Dembski understands very well that assessing fitness at the level of the string smuggles a preconceived objective into the selection processs. Call that front loading or whatever you like. It's quintessential design objective in action.
Comment by Bradford — September 20, 2009 @ 11:43 am
September 20th, 2009 at 11:45 am
Seems fine to me.
ETA
Just to clarify, there is nothing in the weasel program that fixes (or latches) correct letters, they can revert to incorrect letters, but as correct letters are the criterion for choosing the parent, they tend to remain correct.
Comment by Alan Fox — September 20, 2009 @ 11:45 am
September 20th, 2009 at 11:56 am
If it is accurate Alan then Dembski's smuggling charges stick. The selected items are individual letters in fact. But the program assesses fitness at the string level. Not unlike assigning selection to individual genetic mutations affecting amino acids and assessing fitness based on a targeted and currently non-existent protein. A good way to approach front loading and design.
Comment by Bradford — September 20, 2009 @ 11:56 am
September 20th, 2009 at 12:56 pm
Here is the sniveling Weasel I typed out in C:
#include [stdio.h]
#include [cstdlib.h]
#include [time.h]
#define TARGET "METHINKS ITS LIKE A WEASEL"
#define ALPHABET "ABCDEFGHIJKLMNOPQRSTUVWXYZ "
#define TARGET_SIZE 27
#define ALPHA_SIZE 27
char* mutate(char*, char*);
int fitness(char *);
int locked(char *curPtr);
int randInt(int, int);
int main(void) {
char current_string[TARGET_SIZE];
char *curPtr = current_string;
int gen_count = 0;
do {
int i;
char next_string[TARGET_SIZE];
char* nString = next_string;
char buffer[TARGET_SIZE]; char* bPtr = buffer;
nString = mutate(curPtr, bPtr);
if(fitness(nString) >= fitness(curPtr)) { //tests fitness of current offspring and parent
for(i=0;i<TARGET_SIZE;i++) {
*(curPtr + i) = *(nString + i);
}
}
puts(curPtr);
gen_count++;
}while(locked(curPtr) != 1);
printf("%d generations", gen_count);
return 0;
}
int locked(char *curPtr) {
int i;int count = 0; int state = 0;
for(i=0;i<TARGET_SIZE;i++) {
if(*(curPtr + i) == TARGET[i]) {
count++;
}
}
if(count == 27) {
state = 1;
}
return state;
}
char* mutate(char* cPtr, char* bPtr) {
int i;
for(i=0;i<TARGET_SIZE;i++) {
*(bPtr + i) = *(cPtr + i);
}
bPtr[randInt(0, (TARGET_SIZE - 1))] = ALPHABET[randInt(0,(ALPHA_SIZE - 1))];
return bPtr;
}
int randInt(int min, int max) {
static int kState = 0;
int i;
if(kState == 0) {
srand(time(NULL));
kState = 1;
}
i = (rand() % (max - min + 1) + min);
return i;
}
int fitness(char *cString) {
int i;int count = 0;
for(i=0;i<TARGET_SIZE;i++) {
if(*(cString + i) == TARGET[i]) {
count++;
}
}
return count;
}
Comment by computerist — September 20, 2009 @ 12:56 pm
September 20th, 2009 at 1:24 pm
1) Begin with a random string the same length as the target.
YES
2) Using the user set values for offspring, create an array of N progeny strings that are clones of the parent.
YES
3) For each child, mutate letters using the number set by the user for “Mutation”.
YES
4) Pass each child string into a fitness function, comparing them against the target string. Using the distance from that target (checking each letter for correctness at that position), assign values to the strings. Retain the string with the lowest error value.
YES: Here we compare mutated child fitness with previous child fitness (previous child is regarded as the parent). The one closer to the target string goes on to the next generation.
5) Repeat, using the best string as the new parent, until the best string has no errors.
YES
We should distinguish between creating function and selecting function. If I were to categorize Weasel it would be the latter.
Comment by computerist — September 20, 2009 @ 1:24 pm
September 20th, 2009 at 2:04 pm
Accumulation of random mutations leads to a loss of function and inevitably extinction. Without a target, decreasing errors is the last thing that I'd assume Weasel was doing when we look at it from the POV of creating new function, new body plans, it's a pathetic model overall.
If Darwinists fancy it up as anything more then you know they are bluffing:
"Cumulative Selection produces FCSI and IC systems etc…"
Comment by computerist — September 20, 2009 @ 2:04 pm
September 20th, 2009 at 3:00 pm
We got that point Alan. However, what is in dispute, and has not been fully explained by anyone from your side is how “cumulative selection” worked in a pre-biotic environment. All you seem to be arguing is that somehow it did work. Remember this discussion started back on “The Plausibility of the RNA World” thread, where the question of cumulative selection acting upon molecules in an a-biotic (or pre-biotic) environment is relevant.
For example, I have mapped the phrase, IN THE BEGINNING WAS THE CODE, onto the following RNA sequence:
CAAGUAGGGAGUUGAUAAGGGAUAUAAUCACAAGUAGUACAAGUAUCA
GGGUCUAAAACUGGGAGUUGAUAAGGGACAGCAAGAUAA
Let the phrase, IN THE BEGINNING WAS THE CODE, be a stand in for the first functional strand of RNA that leads to a key complexifying replication that starts us down long road of the evolution of life that eventually leads to us. At present I am quite sure that no one has a clue what that first functional sequence is.
The problem is, that in the chemical world, words and letters don’t have intrinsic meaning. The sequence GIBN HNTE WTIGEI ODAS HNE NCE doesn’t have a clue how to arrange or rearrange itself into a meaningful sequence. In the same way neither do nucleotides AUCG for RNA (or the amino acids for protein) have a clue how to arrange themselves into functional sequences. In the a-biotic environment there is no known selection mechanism to “guide” the process. The analogy I have used above coding an English phrase into an RNA is in principle no different the encoding a functionally useful sequence.
And even if such a mechanism, like a primitive computer, existed, it itself is mindless. So what is it? Where did it come from? From where did it get its rules? You can, give an airy wave of the hand and claim that ‘evolution did it’ but you haven’t explained how evolution got that capacity in the first place. Where is the evidence the codons of RNA or DNA (or the amino acids of protein) starting alone in a primitive hostile environment have a capacity by themselves to create meaningful sequences that can evolve into anything?
Paul Davies summarizes the problem succinctly when he writes: “life is more than just complex chemical reactions. The cell is also an information storing, processing and replicating system. We need to explain the origin of this information, and the way in which this information processing machinery came to exist…The problem of how meaningful or semantic information can emerge spontaneously from a collection of mindless molecules subject to blind and purposeless forces presents a deep conceptual challenge.”
Once again we fully accept that WEASEL is an analogy “simply demonstrate the power of cumulative selection“. Unfortunately, it is difficult to see, at least for the present, how cumulative selection could play any role to explain the origin of life.
Comment by JOHN_A_DESIGNER — September 20, 2009 @ 3:00 pm
September 20th, 2009 at 3:41 pm
computerist:
Kill that weasel. Flog it. Stomp on it. Show it no mercy.
Comment by Bradford — September 20, 2009 @ 3:41 pm
September 20th, 2009 at 5:47 pm
Hehe…now that the Weasel is dead the time has come for a replacement
Comment by computerist — September 20, 2009 @ 5:47 pm
September 20th, 2009 at 6:10 pm
JOHN_A_DESIGNER wrote:
I don't know whether to laugh or cry. After weeks of titanic mental efforts the misconception about Weasel appears to be dead—only to be replaced with another silly idea.
Theory of evolution is not about the origin of life, it's about the origin of species. The first
paragraphsentence of the Wikipedia article Evolution mentions that:Comment by olegt — September 20, 2009 @ 6:10 pm
September 20th, 2009 at 6:23 pm
Cumulative selection requires an existing mechanism. Perhaps JAD was unaware that the programmer's hand is the weasel mechanism.
Comment by Bradford — September 20, 2009 @ 6:23 pm
September 20th, 2009 at 6:52 pm
computerist:
I have a reverse weasel concept. Since weasel is dead it's time to skin and dress the animal. After that we start with a string. Let's try Patrick Henry:
Give me liberty or give me death.
This represents a functional nucleotide sequence. Could be the amino acids in a catalytic enzyme needed for generating glucose. An ancient pathway. Could be something else. What matters is that we start with function but this function is early in the game. Too early for error detection and repair functions to have come about. Mutate it with a set of target proteins in mind. Those targets would be chosen based on homology. Of course we need an evolved function in the end.
The question is does the evolved function arise prior to the functional demise of: Give me liberty or give me death? We could have some alternative strings representing alternate functional enzymes e.g:
Give me 12beets or give me abeer.
Do we get evolved function or decay. Which comes first the drowned body or the life boat?
Comment by Bradford — September 20, 2009 @ 6:52 pm
September 20th, 2009 at 7:42 pm
Okay, you would be willing then to concede that cumulative selection plays no role whatsoever as far as the origin of life is concerned? So then how do you explain the origin of life? Is it a miracle, a highly improbable event, or something that can be investigated by science?
Comment by JOHN_A_DESIGNER — September 20, 2009 @ 7:42 pm
September 20th, 2009 at 7:58 pm
John,
The problem of the origin of life is work in progress. At the moment we don't have a scientific theory that addresses that, only some hunches. So I can't answer your question either positively or negatively. It makes no sense to do so at this point.
Comment by olegt — September 20, 2009 @ 7:58 pm
September 20th, 2009 at 8:11 pm
So when adaptation occurs without speciation, that's not evolution?
Comment by Guts — September 20, 2009 @ 8:11 pm
September 20th, 2009 at 8:34 pm
That's evolution, Guts. However, at that stage life has already arisen.
Comment by olegt — September 20, 2009 @ 8:34 pm
September 20th, 2009 at 9:12 pm
Interesting, Bradford.
Comment by computerist — September 20, 2009 @ 9:12 pm
September 21st, 2009 at 12:03 am
Then that definition of evolution is inadequate.
That's a huge can of worms but that's fine. With the risk of derailing this thread further, I'll just say though, that whatever happened at the origin of life, it probably effected it's evolution. If so, it is certainly not 100% separable.
Comment by Guts — September 21, 2009 @ 12:03 am
September 21st, 2009 at 2:07 am
Olegt:
While it is true that Dawkins, at least in his earlier writings, seems to steer clear of extending cumulative selection when it came to OoL, he is hardly the only authority, if indeed he is an authority, on the subject.
Paul Davies in his book, The 5th Miracle, writes that, “Advocates of the RNA-world theory suppose that a soup containing RNA molecules can evolve by a type of Darwinian process. Normally, Darwinism is associated with organisms like cells, but in principle all it needs is replication, variation, and selection. These can occur even at the molecular level, and biochemists use the terms 'molecular evolution' or 'molecular Darwinism' to describe what happens. It is a moot point whether we should define as living anything that can evolve in a Darwinian manner. If so, then perhaps RNA molecules (in a suitable chemical environment) could already be considered living things.”
In other words, according to at least some advocates of the RNA-world, like Manfred Eigen, RNA should be capable of natural and therefore cumulative selection.
Comment by JOHN_A_DESIGNER — September 21, 2009 @ 2:07 am
September 21st, 2009 at 2:47 am
Rubbish. In real life, information is provided by the environment. Niche = target string. Of course, as shouldn't need to be pointed out, the real environment is dynamic and the target string is fixed, which is why Weasel, as Dawkins and everyone else agrees, is not a good model for evolution
Variation occurs in the genotype and selection acts on the phenotype. Front-loading has zero evidence for it as a theory but you are welcome to suggest a way of finding some.
Comment by Alan Fox — September 21, 2009 @ 2:47 am
September 21st, 2009 at 2:52 am
Computerist:
Are you going to eventually be the first person to show how to calculate FCSI 'or even tell us what it is?)?
Comment by Alan Fox — September 21, 2009 @ 2:52 am
September 21st, 2009 at 2:55 am
JohnADesigner:
As Oleg points out you are confused. Darwin's theory was "Origin of Species" not "Origin of Life". Evolutionary theory attempts to explain the diversification of life from the point of the first replicators. You are getting mixed up with the theory of abiogenesis.
Comment by Alan Fox — September 21, 2009 @ 2:55 am
September 21st, 2009 at 3:00 am
Computerist:
Well, Dawkins came up with his biomorphs in the same book. Any thoughts on biomorphs? Any thoughts on how to calculate FCSI?
Comment by Alan Fox — September 21, 2009 @ 3:00 am
September 21st, 2009 at 3:12 am
Guts:
Why not start a new thread on abiogenesis theories then, if you think this is an issue that advances ID. I have sometimes wondered if Behe would have made more headway attacking current abiogenesis theories rather than talking about mouse traps. Is "effected" a typo, as "origin of life carried out evolution" doesn't make much sense.
Of course, unless you are Sherlock Holmes, pointing out weaknesses in theory A does not confirm theory B. Some theory of ID that can be falsified would be a step in the right direction. I see plenty here and elsewhere about how Darwinian evolution is such a bad idea but I don't see much about ID theories that correlate with observed reality.
Comment by Alan Fox — September 21, 2009 @ 3:12 am
September 21st, 2009 at 8:14 am
Double rubbish. The environment does not provide information. It provides a context within which randomly generated outcomes acquire their probability of sticking. The target string transforms that probability onto a certainty. There is no contingency and no variations possible with respect to the unltimate outcome. It's 100% certain down to the dotted i in thinks. Weasel is a good model for directed outcomes.
Comment by Bradford — September 21, 2009 @ 8:14 am
September 21st, 2009 at 8:36 am
In other words JAD, the pruning of randomly generated options awaits an identifiable mechanism about which you are to place complete faith in the ability of the empirical gods to eventually locate. Once they do so you are invited to revere our idol. In the meantime keep the faith.
Comment by Bradford — September 21, 2009 @ 8:36 am
September 21st, 2009 at 9:06 am
Alan Fox:
If you use a realistic mutation rate and a realistic population size a partitioned search occurs.
The way Dawkins describes cumulative selection and illustrates it in "The Blind Watchmaker" the inference to latching is easily made.
Comment by ID guy — September 21, 2009 @ 9:06 am
September 21st, 2009 at 9:15 am
The origin of life is directly related to its evolution.
IOW "how" life arose is very important to its evolution.
Comment by ID guy — September 21, 2009 @ 9:15 am
September 21st, 2009 at 11:17 am
Alan:
Not sure what you're quoting, I never said that. Did you just make that up? What I said was whatever happened at the origin of life effected it's evolution. As in, produced results, consequences, influenced, etc.
Alan:
Alan, time and time again on this blog evidence for front-loading has been presented, even in this thread, it has been demonstrated to you how it can make predictions. Either respond point by point to each and every demonstration or stop making that ridiculous claim. The next time you make that claim it will be an outright lie.
Comment by Guts — September 21, 2009 @ 11:17 am
September 21st, 2009 at 11:29 am
Olegt and Alan, what would you agree or disagree with from the following passage?
Comment by JOHN_A_DESIGNER — September 21, 2009 @ 11:29 am
September 21st, 2009 at 11:29 am
Alan Fox:
Maybe, maybe not.
However it does show that A has weaknesses.
And if there are people claiming A doesn't have any weaknesses, then they have spainin' to do.
Seeing that ID can be tested I would say it can also be falsified.
As a matter of fact- abiogenesis.
Living organisms arising via chance and necessity would refute all of ID's claims.
YOUR reality?
What does correlate with that?
Comment by ID guy — September 21, 2009 @ 11:29 am
September 21st, 2009 at 11:54 am
You two fight it out and, when you come to an agreement, get back to us.
Comment by Daniel Smith — September 21, 2009 @ 11:54 am
September 21st, 2009 at 12:01 pm
Alan and I already agree. I agree with him that Weasel isn't an experiment that scientifically proves something. And he agrees with me that Weasal proves a point, namely that ID's fundamental claim is false because random processes can create information.
Comment by don provan — September 21, 2009 @ 12:01 pm
September 21st, 2009 at 12:05 pm
The difference is semantic. Don is saying weasel disproves (in the sense of falsifies) ID's "random processes can't create information". I am saying Weasel does not prove anything about evolution. I wouldn't have thought of Weasel falsifying any ID claim, as ID hasn't got off first base, as far as I am concerned.
Hope that helps, Dan
Regards
Tool of Satan
Comment by Alan Fox — September 21, 2009 @ 12:05 pm
September 21st, 2009 at 12:06 pm
Guts:
Let me rephrase. I have seen absolutely no scientific evidence for front-loading anywhere on this thread, this blog or anywhere else. I must not be very observant so maybe you could point me in the right direction with a link.
Comment by Alan Fox — September 21, 2009 @ 12:06 pm
September 21st, 2009 at 12:14 pm
We must agree to disagree, then.
The environment is the non-random element in natural selection. It is dynamic and chaotic but deletrious mutations will tend to disappear from the gene pool and mutations that improve fitness will tend to spread.
As everyone agrees, not a realistic model of evolution.
Well, you would know more about that than me.
Comment by Alan Fox — September 21, 2009 @ 12:14 pm
September 21st, 2009 at 12:19 pm
Alan, only nine more months left to confirm your prediction that ID will disappear by mid 2010. Yet here you are in late 2009 devoting precious hours to vigorously defending Richard Dawkins from a recently published ID paper by Dembski and Marks!
[Here's a little ironic truth that's so obvious that you probably can't see it, Alan. ID will never disappear as long as there are people like you around to keep it in the public perspective. And there will always be people like you.]
Comment by chunkdz — September 21, 2009 @ 12:19 pm
September 21st, 2009 at 12:27 pm
In the real world, no maybe about it.
And if there are people claiming A doesn't have any weaknesses, then they have spainin' to do.Evolutionary theory is evolving all the time as new evidence accumulates. Still doesn't alter the fact that perceived weaknesses in ToE do not eliminate the need for an any alternative theory to be well, an alternative theory.
Should I parse this as if science comes up with a passable abiogenesis theory ID will be refuted? I am not so sure, but a workable abiogenesis theory may never turn up, I think we can look at ID theoris on their own merits once we have some.
I don't own reality. It's for everyone that wants it!
Comment by Alan Fox — September 21, 2009 @ 12:27 pm
September 21st, 2009 at 12:53 pm
Lets hear what John A. Davison has to say on the subject of Weasel:
Comment by computerist — September 21, 2009 @ 12:53 pm
September 21st, 2009 at 12:53 pm
lol Alan, you kind of remind me of the black knight from Monty Python. Again, if you don't think that any of the demonstrations constitutes scientific evidence, feel free to respond point by point with reference to the peer reviewed literature. Perhaps you can start with these .
Comment by Guts — September 21, 2009 @ 12:53 pm
September 21st, 2009 at 1:11 pm
Alan Fox:
Are there any non-telic theories we can take a look at to see what is required?
Comment by ID guy — September 21, 2009 @ 1:11 pm
September 21st, 2009 at 1:18 pm
Alan Fox:
It is?
Environments change.
What was good for the parents does not translate into good for the kids.
Comment by ID guy — September 21, 2009 @ 1:18 pm
September 21st, 2009 at 1:26 pm
You're right. Why am I wasting my time? No good reason I can see.
Thank goodness!
Comment by Alan Fox — September 21, 2009 @ 1:26 pm
September 21st, 2009 at 1:35 pm
Guts,
You'll have to reimburse me for replacing irony meters that just blew up.
Your link supposedly pointing to scientific evidence for design goes to the blog of one Mike Gene. As far as I know, Mike Gene does not that ID is science.
And that black knight reference? ID hasn't made any advances, badly lost in court, and has failed to penetrate academia. The Wedge strategy doesn't seem to be working.
Comment by olegt — September 21, 2009 @ 1:35 pm
September 21st, 2009 at 1:35 pm
Oh stop making excuses and respond. It should be easy if it's not actually scientific evidence right? You won't respond because you can't .
The fact of the matter is that the accumulated evidence thus far makes front-loading what Mike Gene calls "proto-science", that simply means that it is not science yet , but the evidence that he accumulates is done through observations, predictions, the scientific method, it is , in and of itself, scientific evidence.
Comment by Guts — September 21, 2009 @ 1:35 pm
September 21st, 2009 at 1:37 pm
Indeed. Why not ask that his posting privileges be restored so we can hear from him. I doubt John has any idea about Weasel; he is a complete computer illiterate!!!
Comment by Alan Fox — September 21, 2009 @ 1:37 pm
September 21st, 2009 at 1:42 pm
Guts,
Maybe it's the IDers that should stop making excuses and present their evidence in the scientific literature. That's the right venue for science. Personal blogs, not so much.
Comment by olegt — September 21, 2009 @ 1:42 pm
September 21st, 2009 at 1:44 pm
Dembski has emailed Wesley Elsberry!!!
Wes has forwarded it to Dawkins. Watch this space!!!
Comment by Alan Fox — September 21, 2009 @ 1:44 pm
September 21st, 2009 at 1:45 pm
Thats just another excuse. IDers are at least trying to, present their arguments in the scientific literature. As far as the DM goes, there's a time and place for everything, in the meantime, everyone can look at the evidence for themselves. From now on, whenever one of you says there is no evidence for front-loading, it will simply be an outright lie, because you have just been corrected.
Comment by Guts — September 21, 2009 @ 1:45 pm
September 21st, 2009 at 1:56 pm
From the link supplied by Alan Fox:
Much is made of the allowance for base (letter) mutations away from the target specification which would occur when the string as a whole better matches target fitness. Reminds me of a fixed boxing event where the boxer, slated to lose in advance of the fight, is allowed to land some real blows against his opponent before he takes a dive. In the end it concludes according to the preconceived purpose. Design by way of directed outcome.
Comment by Bradford — September 21, 2009 @ 1:56 pm
September 21st, 2009 at 2:08 pm
olegt:
It's already there.
You just refuse to see it as such.
Also ID didn't lose in Court. Some strawman lost because the people pushing it a) didn't know what they were doing and b) did it out of religious motivations.
But anyway, what's the alternative to ID again?
What does it have for support in peer-reviewed literature?
Comment by ID guy — September 21, 2009 @ 2:08 pm
September 21st, 2009 at 2:13 pm
While ID critics claim they see no evidence for nature front loading an evolutionary process, in fact, they engage in conceptual front loading by acknowledging a distinction between the origin of life and evolution. The biological mechanisms identified as enabling evolution are themselves posited at the outset of an evolutionary process. They are not empirically accounted for. Yet their front loading is a prerequiste for biological change. Maintaining that there is no evidence for front loading not only discounts deep homology arguments, it of necessity makes assumptions about an imagined process that engendered self-replicating capacity.
Comment by Bradford — September 21, 2009 @ 2:13 pm
September 21st, 2009 at 3:27 pm
Alan:
Not necessarily, genes can be the direct targets of selection, check out the t allele of the housemouse as an example.
Comment by Guts — September 21, 2009 @ 3:27 pm
September 21st, 2009 at 4:21 pm
Jody Hey at Rutgers has a good model.
What is not being modeled correctly by most models is the cost of Purifying Selection.
If a parent has on average a few offspring and each offspring has on average a few novel bad mutations, there is no way natural seleciton can lead to inevitable rise in complexity. Deterioration is guaranteed. There is no way to kill off enough offspring to clean out the genome given this rate. Nobel Prize winner HJ Muller (of Muller's ratchet fame, dubbed by Felsenstein) pioneered this important question. This is genetic entropy. I don't know why such an obvious result is ignored.
Loss of complexity is the more probable outcome by far, not the increase.
This hypothesis can be empirically tested in principle. Simply count the number of unique proteins going to extinction each year versus the number of novel ones. Crude estimates favor the hypothesis of deteriorating complexity, not increasing complexity.
This should be no surprise. Basic population biology would suggest deterioration and degeneration, not improvment and increase in functional complexity.
Papers and videos are on the way to explain more, maybe February 2010….depends on how well the team can push their products out into the public sphere….
Comment by Salvador T. Cordova — September 21, 2009 @ 4:21 pm
September 22nd, 2009 at 8:15 am
Hi, John!
*waves*
Ah c'mon mods! How bad can it be?
Comment by Alan Fox — September 22, 2009 @ 8:15 am
September 22nd, 2009 at 8:32 am
Glad to provide you with amusement! My favourite MP
Re your link to Mike Gene's blog. I had a look, for instance at the piece about mitochondria, and I couldn't see much beyond "how else can you explain it except by front loading?". Sherlock Holmes reasoning is not scientific and does not allow for unknown or overlooked explanations.
Comment by Alan Fox — September 22, 2009 @ 8:32 am
September 22nd, 2009 at 10:45 am
That's not a response, that's just a handwave. A very bad one at that. There is in fact, nothing in that essay that even comes close to saying "how else can you explain it".
Comment by Guts — September 22, 2009 @ 10:45 am
September 22nd, 2009 at 10:50 am
If a parent has on average a few offspring and each offspring has on average a few novel bad mutations, there is no way natural seleciton can lead to inevitable rise in complexity. Deterioration is guaranteed. Comment by Salvador T. Cordova — September 21, 2009 @ 4:21
Non-sequitur, Salvador T. Cordova. Even if mutations were good or bad absolutely, that says nothing about "complexity."
And the idea of an “inevitable rise in complexity” is not and never has been a stipulation or necessary corollary of the theory of natural selection, as Darwin made plain.
“Basic population biology would suggest deterioration and degeneration, not improvment and increase in functional complexity.”
Quite the opposite, although I can certainly understand why one might get that impression by reading some (maybe even most) popgen theory, based as it is on the statistical dynamics of deleterious alleles.
“This hypothesis [‘loss of complexity’] can be empirically tested in principle… Crude estimates favor the hypothesis…”
I can't wait till February. Cite the research, however crude, please.
Comment by Rock — September 22, 2009 @ 10:50 am
September 22nd, 2009 at 11:34 am
My understanding is that in Darwin's mind "complexity = good = reproductive success".
By Lewontin's time "good = reproductive success" with no real reference to complexity. The discussion by Lewontin is here: Santa Fe 2003 (actually you were the one who found this for me years ago. ). Go to page 25:
Darwin believed the environment would induce creation of complexities like the eye or the wing or the heart.
Earth Faces Catastrophic Loss of Species
Presumably, loss of species implies loss of unique proteins or any other bio-entity. The implication is that at best natural selection is destroying, not creating.
Comment by Salvador T. Cordova — September 22, 2009 @ 11:34 am
September 22nd, 2009 at 1:12 pm
Dawkins can't provide the original of his work so that it can be independently analyzed and critqued.
It appears this is Dick's explanation: "I simply lost the source code." or "I really don't want a world class Electrical Engineer and expert on Evolutionary Algorithms like Robert Marks examining the source code of an amateur like me, Dick the Dawk."
C'mon Dick. Surely you can enlighten us ID rubes with your souce code? I mean, it's not possible ID rubes might actually have some legitimate computer science backgrond to actually critique WEASEL?
Comment by Salvador T. Cordova — September 22, 2009 @ 1:12 pm
September 22nd, 2009 at 1:21 pm
I wonder what the rest of the email said. "They don't look familiar" seems clear enough, Sal. Or are you judging others by your own standards?
Comment by Alan Fox — September 22, 2009 @ 1:21 pm
September 22nd, 2009 at 1:27 pm
Here's a suggestion, Nelson. Take your best example of evidence of front loading and start a thread. This would seem more logical than telling me there is a large body of evidence if only I would look.
Comment by Alan Fox — September 22, 2009 @ 1:27 pm
September 22nd, 2009 at 1:28 pm
lmao, Alan, there are dozens of essays over at Mike's blog, I'm just asking you to actually read them and respond point by point if you can. I will post on Front-loading when (and after) I review Mike's book.
All you're doing right now, and all that you will ever do, is simply ignore the points and keep repeating that you don't see it. You're just trolling.
Comment by Guts — September 22, 2009 @ 1:28 pm
September 22nd, 2009 at 1:32 pm
No worries. Take all the time you need. I prefer not to comment on pre-moderated blogs, by the way. First and foremost, I am very much in favour of genuine free dialogue.
Comment by Alan Fox — September 22, 2009 @ 1:32 pm
September 22nd, 2009 at 1:34 pm
Sal,
Anyone reasonably proficient in computer programing and familiar with Dawkins's original description of Weasel can bang out the code in a couple of hours. It's been done on numerous occasions by different people, just search the web.
It's laughable that Dembski and Marks couldn't shoot straight and replaced Weasel with a partition search, which clearly was not Dawkins's intent. It's doubly funny because the misconception has been pointed out to Dembski several times since 2000 (by Wes Elseberry among others). And to top it off, the Isaac Newton of information theory holds a context asking people to produce "the original weasel."
O boy, Bill sure knows how to crack me up!
Comment by olegt — September 22, 2009 @ 1:34 pm
September 22nd, 2009 at 2:08 pm
Dawkins can simply say he no longer has the source code.
If he has lost the code, he can affirm what is an accurate representation of what he really intends by providing source code or saying "so and so has a sufficiently reasonable facsimile".
Can Wes prove his assertion because Wes has a copy of the original Weasel? Does anyone on the planet have an original copy? Does Dawkins?
Isn't an important part of science reproducibility? Dawkins could surely alleviate any ambiguity in interpreting his work by providing source code.
Even in the computer profession, it is very easy to mis-interpret a narrative description of how a piece of software functions. The final description of how software works is the source code.
Jody Hey at Rutgers provides source code to the public. The Avida group of Lenski and Adami provide source code.
Even lil' ole me has provided source code for my evolutionary algorithm (see : GA.C).
And Dave Thomas properly analyzed my "remarkable algorithm" here:
Cordova's Algorithm is Remarkable.
I guess Dawkins lost the code. In such case, he could come forward and say what piece of software is a legitimate reconstruction of his work. If he cares that much about the arguments he made on TV and in his book, I don't think it is too much to ask to have him say what software should be used as a standard to represent his ideas.
Independent of the ID/Darwin issue, I don't think this is too much to ask from a scientific and pedagogical standpoint. He could say, "well someone else made the software."
Fine. Be forthright, be accurate. If a respected member of the IEEE like Bob Marks is wanting to critique a piece of software or algorithm, this is not too much to ask.
Comment by Salvador T. Cordova — September 22, 2009 @ 2:08 pm
September 22nd, 2009 at 2:23 pm
I agree Sal.
Comment by computerist — September 22, 2009 @ 2:23 pm
September 22nd, 2009 at 2:28 pm
You're right Rock, complexity stays, specificity is lost.
Comment by computerist — September 22, 2009 @ 2:28 pm
September 22nd, 2009 at 2:31 pm
Someone explain why a partition search is an inappropriate tool by which to demonstrate cumulative selection.
Comment by Bradford — September 22, 2009 @ 2:31 pm
September 22nd, 2009 at 2:39 pm
Alan:
So I take it you're not a big fan of Elsberry's blog. But anyway, feel free to offer your critique here, you can do it in this very thread. To be honest, I think you are the quintessential "paper tiger", completely harmless , no matter how much flailing you do.
Comment by Guts — September 22, 2009 @ 2:39 pm
September 22nd, 2009 at 3:05 pm
My guess is that if the strength of selection is high enough, then a partition search is a reasonable approximation of cumulative selection.
This debate could be settled by Dawkins providing the source code or declaring what program reasonably represents his ideas.
In fairness to Dawkins, I forget some of the things I write as well. However, if the issue is important enough to him, I don't think he should let it go. That is to say, we all want to know how he thinks natural selection actually works, and then we have a baseline to test his hypothesis as to whether this is what really happens in the wild, or to what extent.
Except for things like anti-biotic resitance, extreme environmental changes, natural selection in the wild doesn't happen with very much power.
Good example: how different are members of one species from other members of that same species. A LOT! There is a high degree of variety. We call this polymorphism. Is polymorphism an evidence or counter evidence to natural seleciton. It is a counter evidence. This was the great debate between the quants like Kimura and Crow versus the selectionists. The quants prevailed. Selectionist theory only prevails in non-quantitative imaginings of Dawkins, it doesn't not squre with hard empiricism or quantitative analysis.
The result of Weasel is a moot point since natural selection is a minor player in organic change.
See this example of a one and two-horned rhino. Larry makes a good case that a random walk is a better explanation for the development of two versus one horn:
http://sandwalk.blogspot.com/2...
Comment by Salvador T. Cordova — September 22, 2009 @ 3:05 pm
September 22nd, 2009 at 3:11 pm
Salvador, I've mentioned this before but it is worth repeating. The true test of a "genetic entropy" concept lies with genomes lacking error correction and repair mechanisms. Mainstream thinking holds that such mechanisms evolve and are explained by selection pressure. I find that dubious.
Comment by Bradford — September 22, 2009 @ 3:11 pm
September 22nd, 2009 at 3:16 pm
I'm just a curious observer. Have you got any evidence that supports a front loading theory? Surprise me.
(P. S. Wesley and I disagree on moderation policy, as a matter of fact.)
Comment by Alan Fox — September 22, 2009 @ 3:16 pm
September 22nd, 2009 at 3:17 pm
lol thanks for the entertainment.
Comment by Guts — September 22, 2009 @ 3:17 pm
September 22nd, 2009 at 3:19 pm
Sal writes:
Well, that settles it then!
Comment by Alan Fox — September 22, 2009 @ 3:19 pm
September 22nd, 2009 at 3:20 pm
No problem. How about some evidence that suggests "front loading".
Comment by Alan Fox — September 22, 2009 @ 3:20 pm
September 22nd, 2009 at 3:24 pm
This reaction makes no sense. Humble Salvador acknowledges he is making a guess. Doesn't quite fit the swampie standard for arrogance I suppose.
Comment by Bradford — September 22, 2009 @ 3:24 pm
September 22nd, 2009 at 3:25 pm
Then I think it is not unreasonable for Dawkins to re-bang out the WEASEL.
Comment by Salvador T. Cordova — September 22, 2009 @ 3:25 pm
September 22nd, 2009 at 3:29 pm
I've suggested a minimal genome, minimal function test. Strip a genome till its error detection and genomic repair functions lie crumpled on the ground. Then observe the level at which the organism loses its viability. Repeated tests involving varied minimal genomes should render a range within which required functions are identified as front loaded.
Comment by Bradford — September 22, 2009 @ 3:29 pm
September 22nd, 2009 at 3:34 pm
OK Bradford,
This is, some sort of hypothesis, presumably. Let's do the thought experiment. Explain more clearly what you intend to do to which genome. Then explain what you expect to see and in what way this will differ from a prediction based on current evolutionary theory. And can you clarify whether front loading is a theory that posits genetic information being present in the genome ready for future use.
Comment by Alan Fox — September 22, 2009 @ 3:34 pm
September 22nd, 2009 at 3:37 pm
Agreed. It is the classic chicken and egg problem, especially at the molecular level where there is a lot of thermal and quantum noise. It's like building a factory in the middle of a tornado, and the tornado never ends.
The elegant XVIVO pictures give the impression that things move in an orderly fashion. It is more subtle than that.
Things move in a purposeful way despite all the Brownian motion tornados. They are able to self-repair from all the buffeting ect. because of the pre-existing error correction. How does a cell function without self-repair, much less evolve?
Selection can actually select against the evolution of novel self-repair since it's pre-cursors are metabolically expensive. The issue of metabolic efficiency over complexity was painfully evident in the Sol Spiegelman experiment. See: Spiegelman's Monster.
Comment by Salvador T. Cordova — September 22, 2009 @ 3:37 pm
September 22nd, 2009 at 3:39 pm
Why should he bother. Weasel served its purpose. There have been many better programs produced since. Are not Dembski and Marks capable of debunking the later examples? What about Dawkins' biomorphs?
Comment by Alan Fox — September 22, 2009 @ 3:39 pm
September 22nd, 2009 at 3:44 pm
Alan:
Looks like you have short term memory since I already provided it. So to remind you I've changed your name link.
Comment by Guts — September 22, 2009 @ 3:44 pm
September 22nd, 2009 at 3:45 pm
Alan Fox has been replaced by:
http://designmatrix.wordpress....
Comment by Bradford — September 22, 2009 @ 3:45 pm
September 22nd, 2009 at 3:46 pm
Alan Fox:
A number of such studies are a matter of record. The focus has been minimal gene numbers. Function seems more pertinent but the idea is to correlate minimal function to viability.
What I expect to see is an emerging boundary separating biological viability from dysfunction. Current predictions indicate a steady progression of accumulated functions with no requisite base level aside from say a self-replicating molecule.
That's the idea.
Comment by Bradford — September 22, 2009 @ 3:46 pm
September 22nd, 2009 at 3:48 pm
A link to one would be most helpful.
Are we doing our thought experiment on a living organism? Your language is a little dense for me. Are you talking about DNA sequences?
Comment by Alan Fox — September 22, 2009 @ 3:48 pm
September 22nd, 2009 at 3:51 pm
Bradford,
If you don't want to engage in free discussion, just ban me. No big deal. What's with linking me to Mike's blog? If you don't want people looking at my old blog, just disable the link.
Comment by Alan Fox — September 22, 2009 @ 3:51 pm
September 22nd, 2009 at 3:54 pm
I did it, since all you're doing is ignoring it and repeating your request for evidence.
Comment by Guts — September 22, 2009 @ 3:54 pm
September 22nd, 2009 at 3:57 pm
Because I and lots of other people bought his book and want to know more.
He should bother if he cares about having his ideas properly understood by the people who take time to study and pay for learning his ideas.
Once the IEEE gets involved, this is the big leagues of information technology, not some creationist journal!
If he wants to snub the publications of such a respected organization, he is free to do so.
What makes you think there are not other critiques in the pipeline?
Comment by Salvador T. Cordova — September 22, 2009 @ 3:57 pm
September 22nd, 2009 at 3:58 pm
A blast from the past:
http://telicthoughts.com/a-gen...
Comment by Bradford — September 22, 2009 @ 3:58 pm
September 22nd, 2009 at 4:02 pm
Guts:
Sorry, Bradford.
As the comment had your name on it, I assumed it was you.
Comment by Alan Fox — September 22, 2009 @ 4:02 pm
September 22nd, 2009 at 4:04 pm
Guts,
If you have all this evidence, surely it will be a simple matter to post a link. I haven't seen any evidence that supports front loading on Mike gene's blog. Link to something specific. Surely you would want to shout this from the rooftops.
Comment by Alan Fox — September 22, 2009 @ 4:04 pm
September 22nd, 2009 at 4:05 pm
Thanks for doing this for me
Comment by Guts — September 22, 2009 @ 4:05 pm
September 22nd, 2009 at 4:06 pm
excuse typo
Comment by Alan Fox — September 22, 2009 @ 4:06 pm
September 22nd, 2009 at 4:06 pm
Past experience!
Comment by Alan Fox — September 22, 2009 @ 4:06 pm
September 22nd, 2009 at 4:16 pm
Not that I expected anything else, but really, guts! Not one little piece of favourite evidence? Not just a link to something specific?
Comment by Alan Fox — September 22, 2009 @ 4:16 pm
September 22nd, 2009 at 4:25 pm
The exchange between Sal and Mesk was amusing, Bradford, but I assume there was some other point you intended, wasn't there?
Comment by Alan Fox — September 22, 2009 @ 4:25 pm
September 22nd, 2009 at 4:48 pm
John says "He is a kind of house punching bag for the management."
You're right, John. But arguing about the merits of evolutionary theory doesn't affect the merits of any competing theory. I would have thought that anyone with evidence that supports their claim would want to explain about it and why it is so convincing. I would expect some real enthusiasm for their ideas. I just don't see it, like the emperor's new clothes, it doesn't seem to be there.
Comment by Alan Fox — September 22, 2009 @ 4:48 pm
September 22nd, 2009 at 5:01 pm
Alan, you are aware that mainstreamers are closet front loaders are you not?
Comment by Bradford — September 22, 2009 @ 5:01 pm
September 23rd, 2009 at 10:14 am
With respect to the cumulaive seleciton versus partitioning, this all may be a moot point.
Consider if:
and
then it follows
In like manner:
Cumulative selection is less probable to succeed than partitioning. Partitioning is less probable to succeed if the partioning if it is not front loaded by intelligence, and without intelligent front loading is no better than random search.
Ergo, cumulative selection, on average is no better than random search unless the selection pressure is front loaded by intelligence.
Thus, Wes, even if his criticism is correct, is arguing a moot point, ergo Dembski and Marks have prevailed over Dawkins.
Is my characterization incorrect?
Recall, I have argued, evolutionary algorithms are known to be no better than average than other algorithms in solving certain problems like the login/password problem, or comparable code breaking problem. This is a very similar problem to selection solving Irreducibly Complex systems.
Whether cumulative selection or partitioning seleciton is used in trying solve the login/password is a moot point, especially since partitioning would be presumed to be more effective than cumulative selection.
If anything, Marks was being generous to Dawkins in presuming partitioning. I don't think Wes is helping his side except to obfuscate the issues (which sort of helps if one doesn't have a defensible position).
That is my humble opinion anyway.
Comment by Salvador T. Cordova — September 23, 2009 @ 10:14 am
September 23rd, 2009 at 10:48 am
WEASEL is an idealization. The more accurate models do not help the Darwinian case. Thus if Marks and Dembski take down the WEASEL, they demonstrate the other models will not be more favorable to the Darwinian case.
if A is less than B, and B is less than C, then A is less than C.
if WEASEL without intelligence is no better than random search on average, and if more realistic models are no better than WEASEL, more realistic models will not fare better than random search.
Ergo, natural selection without intelligence is no better on average than a random search. We see this in the wild. Recall my argument with Mesk about Gammarus Minus. Natural seleciton selection has been seen to select against Irreducible Complexity in the wild, not for it!
Comment by Salvador T. Cordova — September 23, 2009 @ 10:48 am
September 23rd, 2009 at 12:09 pm
I like this comment at UD by someone we all know well, currently posting as:
Comment by Alan Fox — September 23, 2009 @ 12:09 pm
September 23rd, 2009 at 12:30 pm
Is the search as much improved if the intelligently front loaded selection pressure is survival? How about if survival is the selection pressure, but it wasn't front loaded by intelligence?
I guess I'm not sure you've justified "intelligence" here. Your logic supports a pre-existing selection pressure, and I suppose we can agree that "front loaded" is the same as "pre-existing". But I don't see anything in this argument which requires the front loading to be accomplished by an intelligence.
Comment by don provan — September 23, 2009 @ 12:30 pm
September 23rd, 2009 at 12:30 pm
Alan,
You're quote of severesky does nothing to negate the content of the last two postings.
No it is not as absurd:
Comprende?
WEASEL is an example of most ideal, and the only way it works is because of intelligent direction of the selection coefficients. WEASEL without intelligent seleciton is no better on average than random search. If the most ideal will fail, what does that say of realistic models that will not give any better results than the ideal.
Gammarus Minus is a good example of selection in the wild. Same goes for sickle cell anemia and antibiotic resistance due to loss of function and Sol Spiegelman's monster.
Comment by Salvador T. Cordova — September 23, 2009 @ 12:30 pm
September 23rd, 2009 at 12:37 pm
Sal, I was not directly addressing you. I have long given up the idea of a sensible and honest exchange of views with you. The Genetic-ID episode was a lasting disappointment.
Comment by Alan Fox — September 23, 2009 @ 12:37 pm
September 23rd, 2009 at 12:48 pm
Alan wrote:
That pretty much sums it up for you Alan.
Comment by Guts — September 23, 2009 @ 12:48 pm
September 23rd, 2009 at 1:19 pm
Anybody else finding Alan Fox to be utterly boring?
Comment by chunkdz — September 23, 2009 @ 1:19 pm
September 23rd, 2009 at 3:40 pm
Of course.
Comment by Pez — September 23, 2009 @ 3:40 pm
September 23rd, 2009 at 5:42 pm
Lets take a look at one way in which random mutations and natural selection could potentially play a significant role in creating function given a very generalized but specified front-loaded "blueprint" working in parallel. Effectively, the blueprint would give the illusion of natural selection and random mutation acting as the primary cause.
An analogy should do just fine to illustrate the point, lets use the good old fishing net or fishnet. The fishing net effectively not only catches/traps certain objects (fish in this case if the net is intelligently guided) but subsequently dispenses of anything else. This fishnet effectively acts as a filter, basically filtering objects of a certain size. A blueprint of this type would be pretty useless for specifying complex and specified function since its too simplistic, an OR model with virtually zero-qualitative recognition.
A "balloon" shaped net, however, would have a potential sequential storage mechanism (I know, big assumption), capable of storing objects into "memory"; its storage capacity, it would constantly change its internal states, randomly yes but with a narrowed down range of possible input to "listen" for, "absorb" and "explore", and now (having simply defined the spacing of threads – the building blocks of the web net) with the capability to recognize "types" of objects. At this point we could think of the net as filtering in certain "shapes" such as square, triangle, hexagon and filtering out anything else. The current internal states could take the form of building blocks (specified by the net itself), randomly processed until a fitness function "registers" it, consequently the process is neither entirely random nor is it entirely intelligent. Lets look at how we can take this analogy and write up a very simple little program in C (I only add the relevant core concepts):
char stream[26] = "ABCDEFGHIJKLMNOPQRSTUVWXYZ";
struct fishNet {
char input[3];
char output[4]; //lets say we are looking for a "hook" function as per the bacterial flagellum
};
struct filter blueprints[20], *sPtr;
sPtr = &blueprints[0];
//the common objects to specify the HOOK function are HOK, so lets "listen" to those.
sPtr->input[0] = 'H';
sPtr->input[1] = 'O';
sPtr->input[2] = 'K';
char temp;
do {
for(i=0;ioutput[i] = temp;
}
//a print function to print current state (4 character-width)
}
}
while(isLocked(sPtr->output) == 0);
//isLocked() function simply checks the current internal state with the fitness function, if it reached "HOOK" it "registered"
//obviously you need a random number generator if you want to try it out…
*The point is we don't necessarily need to specify the order of HOOK, nor do we need to specify how many exact letters are used, just the common ones.
Now I'm going to be late hehe….gotta go, in a hurry.
Comment by computerist — September 23, 2009 @ 5:42 pm
September 23rd, 2009 at 7:08 pm
My understanding is that in Darwin's mind "complexity = good = reproductive success".
Check your understanding against Darwin’s own words, Salvador T. Cordova. Easy enough to do online.
http://embryology.med.unsw.edu...
Lewontin’s criticism notwithstanding. I bet if one were to ask, most biologists would, like Darwin, take the “metaphorical sense of fitness literally.” But in what sense is “fitness” a “metaphor,” Dr. Lewontin?
“Presumably, loss of species implies loss of unique proteins or any other bio-entity.” Presumably? Implies? When a species goes extinct, there is an obvious loss of its “unique proteins or any other bio-entity.”
“The implication is that at best natural selection is destroying, not creating.”
The press report attributes the cause of increased extinction to “human activity.” Artificial selection; which unlike “natural selection” is creative. In creating an environment we deem fit for ourselves, we are destroying the ecologies to which other creatures have been fitted. There is the obvious loss of etc. as a consequence. Not as a cause, which I thought you were attempting to suggest. But please explain.
Comment by Rock — September 23, 2009 @ 7:08 pm
September 23rd, 2009 at 8:32 pm
Hmmm…some text from my last comment is missing…
Seems the less then operator on the for loop was the culprit.
Comment by computerist — September 23, 2009 @ 8:32 pm
September 24th, 2009 at 1:53 am
Hehe…should have checked my comment before positing it, should be more like this:
Bradford said:
(I thought about your reverse Weasel, Bradford
). Another thing to realize with this example is that error-detection and correction is a consequence of the "blueprint", random processing and the fitness function. The algorithm corrects "unregistered" errors as a function of time. If the system is "registered" via the fitness function, then that requires another error-detection and correction mechanism. Creating the initial function and sustaining it are two entirely distinct processes, although we could say recreating and sustaining are one in the same.
(*sorry for cluttering up the thread)
Comment by computerist — September 24, 2009 @ 1:53 am
September 24th, 2009 at 7:47 am
Assuming from your comment that you are attempting an analogy to evolution, the key is that the first sequence does not exist in isolation, but is part of a community of other sequences. Simple recombination can bring about such evolutionary novelties.
Comment by Zachriel — September 24, 2009 @ 7:47 am
September 24th, 2009 at 10:14 am
Real Natural Selection is what happens in reality, not as biologists hypothesize. What happens in reality is not survival of the fittest, but survival of the luckiest, most of the time. Biologists presume the environment is conducive to making Selection work in the way Darwin envisioned. Real environments often kill off a species before selection (in the way Darwin envisioned) has a chance to work.
The loss of species in the Amazon rain forest are an example of how real selection works in the wild. Nature is under no obligation to accord with the way Darwin thinks the world works. Nature is under no obligation to prevent natural disasters or human intervention from killing off species before they can evolve. Raup's thesis is "Survival of the Luckiest", not survival of the fittest. That seems to have some support from the quants:
If the population is mammalian, this does not bode well for the way Dawkins envisions natural selection. Real selection in the wild for a unique novel trait in mammals is not much different than random. No one really knows how strong the selection coefficients are in the wild: no one really knows the fitness measurements of the nucleotides out there in the wild, they've never been measured to any degree of accuracy. This is the one discipline of science where theories are put forward where fundamental quantities like "fitness" are poorly defined, much less measured in the wild.
Real selection is also MYOPIC, the product of a blind watchmaker. It will therefore select against traits that are immediately burdensome, but which might be useful in the future. Case in point, blind cavefish, sickle cell anemia, cystic fibrosis, etc. ec.
There seem to be systems in biology that appear to have need foresight in their creation, they are not consistent with a myopic kind of evolution.
The problem is biologist characterize Natural Selection as "survival of the fittest", when in reality, the majority of the time it is survival of the luckiest and sometimes the "sickest" (i.e. sickle cell anemia, or other defects).
Also in reality, Natural Selection doesn't weed out mildly bad traits in real populations. This is the problem with Nachman's Paradox. See:
Nachman's U-Paradox
If on average each child has 3 slightly defective mutations, the parents need to reproduce at a rate of 40 kids per set of parents. The number of excess offspring goes up EXPONENTIALLY with higher mutation rates. The math was figured out by Kimura. This is not such bad news for bacterial populations, but it is for mammalian populations like humans.
I and a few associates are putting together a paper and video to clarify the issue so that it's not so obscure. We want lots of debate over the issue. The issue has not been lost upon Oxford geneticists like Bryan Sykes who predicts human extinction in 100,000 years…..
Nachman argues a "fix" is "synergistic epistasis", but I hope to show (with lots of help from Walter ReMine and others), this won't fix the problem.
Bacteria have small genomes and huge reproductive excess. They are more likely to survive, and they are the closest thing to how Darwin envisioned selection to work in removing the under performers. Mammalian populations are not so lucky. Hence, mammals will likely go extinct before bacteria.
The Sea Mink weasel is presumed extinct, but e. eoli with its bacterial flagellum lives on….
Comment by Salvador T. Cordova — September 24, 2009 @ 10:14 am
September 24th, 2009 at 11:51 am
This
preservation of favourable variations and the rejection of injurious variations, I call Natural
Selection. Variations neither useful nor injurious would not be affected by natural selection, and
would be left a fluctuating element, as perhaps we see in the species called polymorphic.
Under Chapter IV Natural Selection, p.247 in the pdf I provided you with, STC.
Hey, I thought I was being helpful in providing the link. Something you could have done for yourself. But I can’t make you read it…
“Luck” favors those well-prepared for it.
http://www.pnas.org/content/10...
But I have to wonder if you aren’t being facetious, Salvador T. Cordova? Does “luck” play such a decisive role in your natural theology? Do you literally believe that God created such an astonishing diversity of life, and then left to chance (luck) their well-being?
Sounds exactly like the kind of natural theology a Neo-Darwinist would embrace. Bad company ruins good morals, STC. I think you should be more selective (so to speak) in the company you keep. LOL
Comment by Rock — September 24, 2009 @ 11:51 am
September 24th, 2009 at 12:13 pm
Again! You left out the most important part of the quote. It's kinda of like you told a joke, but left out the punchline. Twice.
The difference between never becoming fixed, and becoming fixed once in fifty is definitely not random. Sort of like the difference between never missing and hitting your thumb once every fifty times you swing a hammer.
Comment by Zachriel — September 24, 2009 @ 12:13 pm
September 24th, 2009 at 2:37 pm
Huh?
Which means 98% of the time an advantaged trait gets lost.
For evolution of mammalians this isn't to promising.
I didn't say non-random, I said not much better than random.
Random is about 99.9% chance of FAILURE to be fixed. Selectively advantaged in our example implies a 98% chance of FAILURE. Is that sort of improvement that would inspire confidence in a theory?
By the way, a 1% selective advantage of a novel mutation in a human population is exceedingly rare. That 1% figure is a very generous. Any examples of human in the last generation or two that have a 1% selective advantage? If not, then the failure rate approaches 99.9%.
Fixation has a better chance through random drift because there are many more of these mutations without significant selective advantage versus those with.
I reviewed that prior exchange with Raevmo, and if one looks at mammalian populations, versus bacterial, then strength of the argument stands, especially in light of the fact a complex benefeficial (by complex I mean by more than a rew nucleotide) is rare.
Raavmo helped me find the proper qualifier to my original assertion. I miss talking to him. I had nothing to do with his disappearance.
Comment by Salvador T. Cordova — September 24, 2009 @ 2:37 pm
September 24th, 2009 at 2:43 pm
Random: 99.9% chance of elimination
Selectively advantaged: 98.0% chance of elimination
That's not much better than random. In mammallian populations where complex beneficial mutations are rare, this isn't very promising for Darwinian theory.
How many novel beneficial traits (which emerged in the last 100 years) are in the human population with a 1% selective advantage at this very moment?
Comment by Salvador T. Cordova — September 24, 2009 @ 2:43 pm
September 24th, 2009 at 2:45 pm
Rock: "Hey, I thought I was being helpful in providing the link. Something you could have done for yourself. But I can’t make you read it…"
I read it, but I didn't know you had that passage in mind. Thanks for the info! You da man!
In that case, perhaps you and I should hang out more often here.
Comment by Salvador T. Cordova — September 24, 2009 @ 2:45 pm
September 24th, 2009 at 3:03 pm
"During the first seven generations". You just left that out again.
More importantly, if the mutations are selectively neutral, "in an infinitely large population, eventually all the neutral mutations will be lost." All of them.
On the other hand, if there is a 1% selective advantage, there is a 1 in 50 chance of fixation. So, again. Is there is a significant difference between someone's thumb who never misses with the hammer, and someone who hits his thumb every fiftieth time?
Comment by Zachriel — September 24, 2009 @ 3:03 pm
September 24th, 2009 at 3:08 pm
What?
A probability of 0.02 translates into a Falure rate of 98%.
FYI:
1.00 – .02 = 98%
Comment by Salvador T. Cordova — September 24, 2009 @ 3:08 pm
September 24th, 2009 at 3:31 pm
Which is better than 100% failure. That's because no single neutral mutation will fix in an infinite population, but one in fifty with a 1% selective advantage will fix. And that means every fiftieth such mutation will become fixed.
The Ace of Spades only occurs one in fifty-two cards of a standard deck. The probability of the top card of a shuffled deck being the black Ace of Spades may be only one in fifty-two, it is far higher than the probability of seeing a red Queen of Clubs.
If your point is that not all mildly advantageous mutations will fix, you are correct. But your claim was that "Real selection in the wild for a unique novel trait in mammals is not much different than random." Not only do slightly beneficial mutations become fixed at a given rate, but some mutations have higher (or lower) benefits than others.
Comment by Zachriel — September 24, 2009 @ 3:31 pm
September 24th, 2009 at 4:09 pm
Populations are not infinite in size, especially mammalian. My claim was related to mammalian populations.
The majority of "fixed" molecular mutations are not subject to selection (ala Kimura).
What is especially bad is that one does not need to "fix" (as in infuse to every member) the same bad mutation in every individual for bad mutations to degrade an entire population.
Neutrals do not need to be fixed in order to radomize or damage a genome!
Take 1000 individuals with a 4 giga base pair genome each. Give each 1 billion base pairs of random muations that are neutral (if such a thing is possible without killing the population). None of the mutations need to be fixed for the genome to be effectively randomized.
If every member has a bad mutation unique to him and passes it on to his offspring plus more, the damage is done. One individual could have diabetes, another sickle cell anemia, another myopia, etc.. This sorry state of affairs does not get better so long as new bad mutations keep getting introduced with each generation, especially if the bad mutations are only mildly deleterious (versus fatal).
So the situation is not symmetric with respect to good and bad mutations. Good mutations need to be fixed into the population to have lasting good effects, bad mutations do not need to be fixed into the population to have lasting bad effects so long as a supply of bad mutations emerges every generation (and we know this to be the case in human populations). This accords with what we know in engineering: it is easier to destroy than construct….
Thus, unless the fixation rate for positive mutations per genration is sufficiently high, it will not be able to overcome the effects of randomization or bad mutations. The reality of the situation is that for every step forward, there are a hundred backward.
The way Weasel would be more in accord with reality is to model the high frequency of bad mutations (letters deviating from the target phrase) relative to the fixation rate of good mutations.
The proper cumulative result would be that even in the unlikely event "METHINK LIKE A WEASEL" is reached for a season, it doesn't last long before degenerating. Also, a random catastrophe should be modeled every now and then just to illustrate how random events of 1 day can wipe you millions of years of evolution…
Comment by Salvador T. Cordova — September 24, 2009 @ 4:09 pm
September 24th, 2009 at 9:23 pm
While providing supposed support from an analysis of infinite populations that you then misinterpreted. Again.
If the mutations are neutral, then they have no effect on fitness—by definition.
Though rarer, beneficial mutations are much more likely to be fixed, while deleterious mutations are more likely to be eliminated. Yes, nearly-neutral mutations can accumulate in genomes, but natural selection tends to purify genomes over time.
Comment by Zachriel — September 24, 2009 @ 9:23 pm
September 25th, 2009 at 10:38 am
You don't get it! Fixation for a given neutral is even stronger in a finite populations. The baseline of infinite size was a generous assumption for the selectionist side, so the case of a finite population makes the argument I put forward even stronger.
Take a population of 10 individuals. 1 of them has a beneficial mutation with a 1% advantage. Say nine of the ten don't pass on their genes but one with a neutral mutation passes on his genes by random luck and the advantaged individual doesn't pass his genes (say because of a random accident).
By the way, it's wrong to think a single individual can overtake a population of size infinity in finite time. The assumption of infinity isa mathematical idealization to see where things will approach.
But even granting an infinite population, the situation is not symmetric! How many unique traits in an infinite population can have a 1% advantage? It's a bit ridiculous to argue 2000 can have a 1% advantage…consider that we have 2000 individuals with each a 1% advantaged unique trait.
To make the extreme illustration, you can't have a population of 10 individuals with each individual having a 100% advantage. You can't have 10 individual each having a 50% advantage over the rest, so on and so forth.
On the other hand, in an infinite population you can have an infinite number of neutrals!!!!! Thus the fixation rate of neturals is equal to the mutation rate, since we are not tracking any specific netural.
Thus it is conceivable that more neutral traits can be fixed over infinite time and infinite population than selectively advantaged traits if there is not sufficient number of new selectively advantaged traits over time.
A mutation does not have to be positively fit to incur damage: sickle cell anemia, blindness in cave fish, thus in principle even a "neutral" mutation can be damaging to complexity.
Reproductive success does not equal evolution of integrated complexity, and that is what is at issue. Integrated, designed complexity is not necessarily equal to reproductive success.
"Fitness" is poorly defined notion with respect to the evolution of integrated complexity. That is why selection can favor dysfunctional systems.
It depends on reproductive capacity. For small bacterial genomes, yes, for mammalian genomes where there are relatively few offspring, no.
If a human parent confers 100 harmful mutations to his kids on average, how many kids must the parent have to ensure the genome is cleaned out. LOTS!
"Much more likely" is not true in small finite populations. Even in large mammalian populations this is a dubious claim. Case in point. What novel beneficial mutation in the human population has emerged in the last 100 years that looks like it will overtake the human race? Or how about even a nation, or a state, or a city?
Comment by Salvador T. Cordova — September 25, 2009 @ 10:38 am
September 25th, 2009 at 4:28 pm
Nope, natural selection in the wild is not much better than random. See Paleontologist David Raup's book: Bad Genes or Bad Luck.
If a complex tribolite gets killed by a natural disaster, does that mean the critter is less fit? If a creationist kills a Darwinist (as reported here Evolution vs. Creation Ends in Stabbing), does that mean the creationist is more reproductively fit? What constitutes "fit" is in the eye of the beholder. One wonders then how subjectable to empirical inquiry fitness can be if the definition is so unstable.
In the case of antibiotic resistance where bacteria survive because they have defective pumps, is that really a notion of "more fit"? In the case of pesticide resistance because the breathing skin of the insect is dysfunctional, is the creature really "more fit"?
If what defines "fit" is unstable, then how can one claim natural selection is the exact opposite of random?
And given the considerations from molecular genetics, where fixation of a random neutral (if we are talking about some neutral, not a particular neutral) has as much if not more success than an advantaged trait, then how is NS the exact opposite of random? (recall the situation with neutrals is not symmetric with advantaged traits because the sheer number of neutrals can swamp the number of advantaged traits).
Dawkins WEASEL only works under unrealistic (dare we say un-natural) assumptions. Real selection in the wild is unlike the way Dawkins envisions.
WEASEL is an excellent pedagogical tool to elucidate Dawkins misconceptions of evolutionary biology. The quants like Kimura and Michael Lynch and JBS Haldane have a much better grasp of how evolution really works in the wild.
If I had to pick a naturalistic explanation for the appearance of Design, I'd say either it was an artifact of our perception (Nei and Michael Shermer) or some sort of self organization or self evolution (Kaufmann or James Shapiro). Natural Selection would be my last choice. Personally, I think ID is the best choice, but I at least want to acknowledge what seem to be better naturalistic explanations than "natural" selection or Dawkins WEASEL.
Comment by Salvador T. Cordova — September 25, 2009 @ 4:28 pm
September 25th, 2009 at 6:46 pm
According to Social Darwinism, yes.
Comment by chunkdz — September 25, 2009 @ 6:46 pm
September 25th, 2009 at 6:59 pm
You forgot the "unwitting" part, (I hope)!
Comment by Daniel Smith — September 25, 2009 @ 6:59 pm
September 25th, 2009 at 7:04 pm
Alan,
The body of evidence is "biological life". It's the same evidence, there are just many different interpretations. Your bias determines the interpretation you prefer, mine = mine. It's really as simple as that.
Comment by Daniel Smith — September 25, 2009 @ 7:04 pm
September 25th, 2009 at 7:06 pm
Well, you got the "tool" part correct.
Comment by chunkdz — September 25, 2009 @ 7:06 pm
September 25th, 2009 at 10:24 pm
Be specific. What is the probability of fixation? You may want to consider the simpler case of neutral drift, or a more realistic population size.
An infinite number.
Also, there is a significant difference between absolute advantage and relative advantage. A small absoltue advantage in a trait can lead to much a high relative advantage. Consider a species where males compete for harems, winner take all. Even a small difference in strength or agility can mean a particular male's genes predominate in the next generation. Or in a sexually choosy species, where a pretty face can mean the pick of the healthiest mate.
Comment by Zachriel — September 25, 2009 @ 10:24 pm
September 25th, 2009 at 10:50 pm
Zachriel: An infinite number.
Sorry. It's limited by the number of traits in the genome. We might consider the case of an infinite genome, or a trait with unlimited flavors, but assuming a finite genome, then this number is finite.
Comment by Zachriel — September 25, 2009 @ 10:50 pm