When Improbabilities Become Exponentially Improbable

I am sure there is plenty ofphony math out there. John A. Davison

John Davison, "Is not all of mathematics natural?" Theoretically. But in practice, I've found Numerical Analysis particularly unnatural. Or at least awkward. Douglas

Salvador: "I was especially eager to hear what you had to say about the book given your background in science from Cornell." I'm not the person you are pointing to. I hold a PhD in Electronics and Computer science and I have never studied at Cornell. kairos

Kairos, Thank you so much for your informative response. I was especially eager to hear what you had to say about the book given your background in science from Cornell. regards, Salvador scordova

John Davison:

Thanks for the plug PaV but Davison doesn’t have a “theory.” What he has is a new hypothesis which remains in complete accord with everything we know from the experimental laboratory and the fossil record.

Indeed, you're right! It is an hypothesis, and a good one at that. John Davison:

Everything in the world is natural or it wouldn’t be there. All things became natural when the Creator or Creators produced them. Is not all of mathematics natural? Does anyone still think that mathematics was the product of the human mind? Science is nothing but discovery of what has always been before us just waiting to be disclosed.

I couldn't agree with you more, John. The Pope, in Regnesburg, made it apparent that the way he plans to attack the tyranny of the sciences is to point out this very fact; that is, the world has a rational basis. (In physics you would call it Hilbert space.) And all scientific knowledge is possible only because of this rational basis is imposed, rather than being a product of nature itself. Hence, this rational basis implies its own creation, and so, reason leads us to "beingness", which, as the Pope points out and terms it, comes to us as a Christian/Hellenic inheritance. Thus, theology and philosophy--two disciplines that concern themselves with "beingness"--are rightly positioned before the human mind (our rational faculty). Excuse the prolix, but philosophy requires philosophy. PaV

Tom English: I know the feeling, and can sympathize. PaV

kairos and PaV, Sorry not to respond properly. I need desperately to get some work done. Tom English

PaV Everything in the world is natural or it wouldn't be there. All things became natural when the Creator or Creators produced them. Is not all of mathematics natural? Does anyone still think that mathematics was the product of the human mind? Science is nothing but discovery of what has always been before us just waiting to be disclosed. As for evolution, it is FINISHED and has been for a long time. "Here I stand. I can do no otherwise." Martin Luther "A past evolution is undeniable, a present evolution undemonstrable." John A. Davison John A. Davison

#52 Salvador: "Have you had the chance to read Genetic Entropy yet? I would be curious to hear your comments. The issue is taht even if a functional protein formed, what’s the chance it will remain in that state to be any good. Yes, I read it last July (and I also wrote a short comment for Amazon). My opinion (I cite my comment) is that this book provides very good, clear and useful information about the real status of neodarwinian mechanism, mainly concerning the inadequacy of random mutation+selection to both repair genome degradation and macroevolution. Moreover in my opinion it is of particular importance that many nails used for the NDE coffin are provided directly by strict darwinian researchers such as Haldane and Kimura! My main criticism is that the author in some points has forced his (good) arguments to support a specific YEC position (actually the age of the first men). IMHO it would have been better to provide only the scientific arguments with their intrinsic strength. Anyway a very good work in my opinion. Note for #1 reviewer: it is not correct that the author does not consider duplication as a possibility for random evolution. Indeed, in his answers to objections at the end of the book, he explains why gene duplication does not theoretically improve the plausibility of the primary axiom. kairos

Thanks for the plug PaV but Davison doesn't have a "theory." What he has is a new hypothesis which remains in complete accord with everything we know from the experimental laboratory and the fossil record. He also rejects special creation, divine intervention or any other device which denies reproductive continuity. Man is very definitely an animal as this and every other forum demonstrates beyond any reasonable doubt. "A past evolution is undeniable, a present evolution undemonstrable." John A. Davison John A. Davison

Tom English: "Knoll also indicates that between 800 million and 1.3 billion years passed between the appearance of the first eukaryotes and the appearance of eukaryotes clearly related to modern groups. Decide how many base pairs you want the genome to gain. Divide that number by 800 million years." I think all of this number-crunching shows how little we're able to use this information. For example, dividing by 800 million, the 8 billion base pairs we have amounts to 100 base pairs/yr. We could probably come up with all kinds of numbers---but how informative will they really be. "I’ll just give you a teaser: I suspect that in sexual reproduction the ovum contributes a considerable amount of information to the offspring. This is my only point of agreement with Jonathan Wells." I agree that the ovum contributes considerable information--in one of the most hard to read (impenetrable and boring come to mind) book ever written, What Genes Can't Do, Lenny Moss makes the fundamental contribution of cells (ovum) abundantly (and repetitively) clear--but this is no different than panspermia: it simply pushes the problem back. The problem then becomes where did the cell acquire this information? I don't see how this helps. Without much serious thought, I would think two mechanisms are available. One is "front-loading", which simply says that when cells came into existence the original chromosomes had all the information encoded to produce all of the abundant and diverse life we see, with chromosomal rearrangements (a la Davison's theory) and deletions resulting in the genomes we now see. This has a lot going for it. It certainly is tidy. The other mechanism is de novo "creation" of new genomes, at first representing producing kingdoms and phyla, and later producing the major classes we see, with simply chromosomal rearangements producing families of organisms, and adaptation bringing about genera and species and varieties. There are abundant theological problems with this (just ask the Creationists and, from the Catholic side, the Thomists), but this is consistent with, or at least applicable to, the work of "Wisdom"--a personified, but less than divine, power--that we read about in Wisdom literature. Darwinists, of course, would howl at this approach. (But I, of course, howl at what they have to say, so I guess that's a stand-off.) It might be virtually impossible to separate these two mechanisms from each other, OR, from a more naturalistic explanation as well. (God is humble. He stays hidden, and only reveals himself to the humble.) But to try and explain what we know of plants and animals, the fossil record, molecular biology, etc, etc. while invoking Darwinian mechanisms just can't be done. I don't think it will be too long before it all collapses. Well, just some musings. PaV

Kairos, Glad to see you. Have you had the chance to read Genetic Entropy yet? I would be curious to hear your comments. The issue is taht even if a functional protein formed, what's the chance it will remain in that state to be any good. Salvador scordova

For all you chance-worshipping atheists out there, please document a single instance supporting random mutations as creative elements in any evolutionary sequence. While you are at it show me a single sexually reproducing animal that will ever become another species or that ever did in the past. After you have admitted that you can't, I have a few more things for you to concede as nothing more than figments of an overactive collective human imagination. Of course you will never admit that you can't because to do so means that your entire view of the living world is meaningless nonsense. I certainly don't expect anyone here to do what William Bateson did as I have documented both here and in hard copy. I refer you to the side board and my published paper "The blind alley: its significance for evolutionary theory." Neither Mendelian (sexually mediated) genetics nor allelic mutation nor natural selection ever had anything whatsoever to do with creative evolution except to ensure ultimate extinction. At that they have been superlative. "A fanatic is one who can't change his mind and won't change the subject." Winston Churchill Carry on with your fantasies. Don't pay any attention to me. At this point I wouldn't know how to handle it anyway. It is hard to believe isn't it? I love it so! "A past evolution is undeniable, a present evolution undemonstrable." John A. Davison John A. Davison

PaV, "I don’t understand why you’re making the distinction between “random” and “pseudo-random”. Can you predict the output of a rng?" The pseudo- truly is important in practice, though the question of whether anything is "really" random is philosophically deep. Cryptologists have cracked cryptosystems by exploiting predictability of pseudorandom number generators (PRNGs). The response has been to develop cryptographically secure PRNGs. They are harder to predict. "If you want a “real” random number generator, then why not use a Geiger-Counter and a radioactive source" That's what the HotBits web site does for you. See the link at post 17. PaV: "Eukaryotic life appeared over a relatively short period of time. Bacteria dominated up until shortly before the pre-Cambrian. Without bothering to go look at the numbers, off the top of my head, I would say that you have, at most, 3-400 million years to basically come up with the eukaryotic genome, which has billions of bases." Wiki: "Knoll (1992) suggests [the eukaryotes] developed approximately 1.6 - 2.1 billion years ago. Fossils that are clearly related to modern groups start appearing around 800 million years ago." So Knoll indicates that at least 1.4 billion years passed between the appearance of bacteria and the appearance of eukaryotes. The modern eukaryote S. pombe has just 14 million (not billion) base pairs. I take this as an upper bound on the number of base pairs in the first eukaryote. Dividing the latter quantity by the former gives one base pair per hundred years. Knoll also indicates that between 800 million and 1.3 billion years passed between the appearance of the first eukaryotes and the appearance of eukaryotes clearly related to modern groups. Decide how many base pairs you want the genome to gain. Divide that number by 800 million years. "Well, Tom, what’s your proposed mechanism?" Well, it's certainly not viruses from outer space. It's late. I'll just give you a teaser: I suspect that in sexual reproduction the ovum contributes a considerable amount of information to the offspring. This is my only point of agreement with Jonathan Wells. Tom English

#47 "Thus the probability that a random permutation of 100 amino acids represents cytochrome C is 3.8 x 10 ^ -69. That’s a very low probability, but it is much greater than Bill Dembski’s universal probability bound of 10 ^ -120. Furthermore, there are supposedly many sequences of 100 amino acids that represent proteins other than cytochrome C." Let us (only for a moment) accept this computation as plausible. However 10^-69 is a probability enormously low and you cannot save your reasoning by simply observing that it is higher than Bill's UPB. First, this number is 10^-19 times the Borel's UPB (a quite sufficient number to declare something as pretty impossible). As the estimated universe time is about 10^17 seconds, this means that this level of probability can be reached by accumulating a sequence of dependent events, one for each second, with Borel probability during all the universe time. Are you realizing that value of your argument? "Note also that the human-designed protein chignolin has only 10 amino acid residues. This suggests to me that life could have started very small." Instead, the fact that in nature the smaller protein is very much larger suggests to me something completely different :-) kairos

#35 "Evidently you have the Cambrian proliferation of body plans in mind. Due to the fact that gene expression is highly nonlinear (with genetic regulatory networks, protein interaction networks, and inter-cellular communication), small changes in genotype can account for big changes in phenotype. Provided we stick to complexity in the sense of Kolmogorov (algorithmic information, as I specified above), there’s no evidence to back up your claim." Tom you are completely reversing the burden of the proof. Actually there is no evidence at all that non-linearity in gene expression could be useful in finding the right path in phenotype evolution. In (good) science something that hasn't a decent chance to occur is discarded as a plausible explanation. Please prove before that your supposedly magic mechanism is able to get your claim and we'll discuss after. "The notion that larger populations search a fitness landscape better than smaller populations is a common misconception. Under various circumstances, a small population is more likely to escape a local maximum than is a large population." What are you saying? A small population simply provides much fewer tries in the space search and this produces a MUCH LESS capability to find decent solutions. Think to the genetic algorithms. “Almost all fitness functions are irregular in the extreme. That is, they are algorithmically random, and thus they are easy to optimize. Let’s say we uniformly draw a fitness function from the set of all functions from the space of genotypes to the set of all fitness values. Irrespective of the size of the space of genotypes, 916 fitness evaluations yield with 99.99% certainty a genotype that is more fit than 99% of the genotypes." Only when the search space has certain characteristics and aonly at the extreme; but you cannot argue in this sense for the biological search spaces. Simply, in the meantime you must accept that the search space is not reasonably affordable by RM+NS kairos

PaV, Thanks for putting me onto cytochrome C. I am out of my depth here, but I have learned poking around the web that 3.8 x 10 ^ 61 of the 10 ^ 130 sequences of 100 amino acids represent cytochrome C (H. P. Yockey, 1977. On the Information Content of Cytochrome C. Journal of Theoretical Biology, 67: 345-76). Thus the probability that a random permutation of 100 amino acids represents cytochrome C is 3.8 x 10 ^ -69. That's a very low probability, but it is much greater than Bill Dembski's universal probability bound of 10 ^ -120. Furthermore, there are supposedly many sequences of 100 amino acids that represent proteins other than cytochrome C. Note also that the human-designed protein chignolin has only 10 amino acid residues. This suggests to me that life could have started very small. "Would you like to explain where the extra 16 mutations came from?" Data dredging. (Wiki has a disappointing article on the topic.) You might as well ask me to account for the long regions with no mutations. The researcher, Katherine Pollard, had a computer plow through the human and chimp genomes to find regions with substantial differences. She identified 49 "human accelerated regions" (HARs). What makes HAR1 number one is that it is, as Pollard points out, "really an extreme case." http://www.sciam.com/article.cfm?chanID=sa003&articleID=00023D61-9116-14E3-911683414B7F0000 So what I am saying is that if you slide a window over two sequences of 3 billion base pairs looking for anomalous regions, find 49 of them, and then report the very most anomalous, it really is going to look anomalous. All base pairs have been mutated many times in both species in the past six million years. The question is what happened in humans that allowed mutations in HAR1 to be adaptive when they are lethal in chimps and many other species. My uneducated guess is that what ultimately accounts for the high mutability in HAR1 is some prior mutations outside of the region. Tom English

Tom: "Monte Carlo simulations use random numbers. They do not generate them. Your argument was that the only useful randomness in programs was “programmed in.” But all “programmed in” random number sequences are pseudorandom. A “truly” random sequence has to be input to the program — it is not programmed in." I didn't say that the Monte Carlo simulations produced random numbers as output; I said that they produce random numbers, which are used as input. I don't understand why you're making the distinction between "random" and "pseudo-random". Can you predict the output of a rng? If you want a "real" random number generator, then why not use a Geiger-Counter and a radioactive source, and add (integrate) the number of 'hits' over a selected time interval? T.E.: "I genuinely do not know what you are trying to say. If you agree that there were ancestors of humankind living 3.5 billion years ago, then I have, as I intended, been conservative in using the figure. That is, information has accumulated in the genome over more than 3.5 million years." I'm trying to tell you that it is easier to find a way to bacterial life than it is to eukaryotic life. Eukaryotic life appeared over a relatively short period of time. Bacteria dominated up until shortly before the pre-Cambrian. Without bothering to go look at the numbers, off the top of my head, I would say that you have, at most, 3-400 million years to basically come up with the eukaryotic genome, which has billions of bases. I say it's easier to argue to bacteria because you're probably dealing with 100's of millions of years, but the bacterial genome is of the order of 1 million bases. So with eukaryotes, you have 1000 times more work to do, in about twice the time. T.E.: "Also, if you read above, you will see that I told Gil that he really had no way to respond but to attack gradualism and invoke irreducible complexity. You have been johnny on the spot." Well, Tom, what's your proposed mechanism? PaV

Dave, Old? I did patchboard programming in 1974-75. Tom English

Tom English: "It is so very, very easy. First you predicate chance (sans necessity), and then you stipulate that a particular protein structure must emerge, just so, at once. But it is much harder to show that no protein structure whatsoever could ever emerge through evolution." I wouldn't quibble with what you stated. However, you are asking for a proof by negation, which is extremely hard, while at the same time ignoring that it seems as though Darwinists ought to explain how these protein molecules came about in the first place instead of asking someone to prove that every known protein molecule couldn't have emerged. You're also ignoring the fact that Fred Hoyle, Nobel Prize winner, took one look at the calculation I spoke of, and concluded that Darwinism was hokum. Accuse him of being a Creationist? Well, he was an atheist, and in lieu of Darwinism proposed panspermia. "In all data transmission, error correction reduces the probability of error, but it does not reduce it to zero. You can’t beat the Second Law. Think of it this way: What would correct an error in the error correction system? Or think of it this way: Sometimes a fruit fly has an extra leg in place of an antenna." Think of it this way: in an article in the latest issue of Nature, there is a study which identified so-called HAR genes. The authros state there that the HAR1 is a particularly highly-evolved gene in comparison with the chimpanzee genome, meaning that it has 18 mutations instead of the 1 or 2 that would be expected through random mechanisms. Would you like to explain where the extra 16 mutations came from? I'm in a rush. I'll post more later. PaV

There is not a shred of direct evidence that any new information has entered the genome of any living thing during the course of its evolution. It may have of course but if it did where did it come from? There is also not a shred of evidence that bacteria ever evolved into anything but bacteria. The vast majority of bacteria are only varieties of a relatively few basic forms which cannot be transformed even from one to another. To be brutally honest there is no convincing evidence that any phylum ever evolved into any other phylum. They all appeared almost simultaneously including some that became extinct almost immediately. This does not mean that they were separately created however. It means only that early in the evolutionary scenario that the "prescribed" body plans were expressed very early. The history of the fossil record is the story of progressive loss of potential. The first capacity to be lost was that of producing the fundamentally different body plans that characterize the various phyla. Once produced no new ones ever appeared and some disappeared early on. Next to appear were the various classes within the phyla and once they were evident no new ones would ever appear and again some would soon disappear. Next to appear were the orders, followed by the genera and finally the species. Unlike the phyla, most have which have remained, whole orders, families, genera, and species have appeared only to become extinct. The creative capacity has steadily declined until today there is no genus that can be demonstrated to have appeared in the last two million years, during which time untold numbers of species have come and gone. I am now convinced that even no true species are being produced at present. Grasse has suggested the same as I have shown. In support of that is the failure for anyone to demonstrate creative evolution when I have offered the challenge. The present flora and fauna are the terminal products of a finished evolutionary process which will never resume. How can one further refine the structure of the elephant, the lion, the eagle or the horse? The answwer is one can't. They are the ultimate products of an orthogenetic goal-directed process which has terminated with those final immutable products. So also is Homo sapines. I repeat that we are the ultimate product of a planned and completely realized purposeful evolution in which there was never any role for chance whatsoever. Mindlessly to assume otherwise as the Darwinians insist on doing is with foundation. They apparently were simply "born that way." I am unable to offer any other explanation. The entire process was determined from begining to end and chance had no role in it. To imagine otherwise is atheist inspired mysticism and nothing more. There now, I feel somewhat better. Thanks for not listening. As near as I can tell no one ever does. For that reason alone - I love it so! "A past evolution is undeniable, a present evolution undemonstrable." John A. Davison John A. Davison

Tom Sure. The sample rate should be set by the A/D convertor. Usually that's done through an interrupt or DMA controlled by the convertor itself. That way it only records a new sample when a new sample is available. A file produced by the API should be produced in that manner. Preferably you want a 16 bit sample and take only the lowest order bit which should be noise even if it's got a live mike so long as it isn't being driven to ground or saturation. Shift the low bit of consecutive samples into a variable of the size you need and that should do it. The full size samples will have some redundancy in them as cross-talk from regular frequency sources (the switching frequency of switched power supplies for example) will show up in the sample data. That's where the "hum" comes from in old amplifiers turned up too far or those with deteriorating power supplies - the 60-hertz AC line frequency (or 50 hertz if you're across the pond) isn't adequately filtered out of the DC after rectification when electrolytic caps start drying up. It's actually 120 hertz after the rectifier. Modern power supplies are almost all switchers which operate in the high kilohertz range typically above human hearing so the hum from the old days is a relic seldom heard anymore. Now this is making ME feel like a relic from the old days. Thanks a lot. :razz: DaveScot

Dave, Thanks. The reason I have worried about such sources of random numbers is that I know that the people measuring atmospheric noise go to some lengths to clean up the data and make it "more random." You would think data from a random noise source would not need cleaning up, but if you, say, sample at too high a rate, you can get correlated values. Again, thanks. Tom English

Tom English I'm surprised you didn't know that a/d convertors can provide a ready source of high quality random number samples. In the immortal words of Maxwell Smart, "That's the oldest trick in the book". Here's some free source code in Visual C++ for an A/D based rng using the sound API of DirectX to get at the hardware. http://www.cyotec.com/resources/cyorand/ The following source code should show you how to get the raw sample data from a Mac to replace that from the DirectX sound API. http://developer.apple.com/samplecode/RecordAudioToFile/index.html I'm an IBM PC guru. The last Apple programming I did was almost 25 years ago on an Apple IIe. DaveScot

Pav: "Bad calculation. The first form of life is bacterial. It shows up about 3.5 billion years ago. The conditions for life on the early earth didn’t begin until roughly that time. So I’m afraid you don’t have 3.5 billion years to work with." I genuinely do not know what you are trying to say. If you agree that there were ancestors of humankind living 3.5 billion years ago, then I have, as I intended, been conservative in using the figure. That is, information has accumulated in the genome over more than 3.5 million years. When the conditions were right for life in the oceans and in the earth is a matter of debate. "Continual increase over time is not a plausible mechanism for the accumulation of information." If you read what I wrote above, I spoke of the mean gain and the rate of gain. Neither of these implies continual gain. Also, if you read above, you will see that I told Gil that he really had no way to respond but to attack gradualism and invoke irreducible complexity. You have been johnny on the spot. So just to make everyone happy, let's emphasize that, assuming a bacterial ancestor of homo sapiens lived 3.5 billion years ago, and assuming for simplicity that the bacterium had no genetic information, evolution has gained ON AVERAGE 1.4 bits of information per year to reach the human genome. Tom English

PaV: "It’s really quite easy to demonstrate that proteins have very little likelihood of forming by chance." It is so very, very easy. First you predicate chance (sans necessity), and then you stipulate that a particular protein structure must emerge, just so, at once. But it is much harder to show that no protein structure whatsoever could ever emerge through evolution. "Do you have some idea as to how they formed by chance, as in the first cell?" As a researcher in evolutionary computation, I have never had much reason to delve into the orgin of life. And it is funny that you should bring it up now, given that we were talking about evolution. "So there are now better generators than a Monte Carlo simulator. So what. How does that affect my argument?" Monte Carlo simulations use random numbers. They do not generate them. Your argument was that the only useful randomness in programs was "programmed in." But all "programmed in" random number sequences are pseudorandom. A "truly" random sequence has to be input to the program -- it is not programmed in. Tom: "You seem to have problems with the Second Law of Thermodynamics. That is why there must be occasional errors in reproduction." Pav: "And you seem to have problems with the repair systems that biological forms have in place. Next question please." In all data transmission, error correction reduces the probability of error, but it does not reduce it to zero. You can't beat the Second Law. Think of it this way: What would correct an error in the error correction system? Or think of it this way: Sometimes a fruit fly has an extra leg in place of an antenna. Tom English

mike1962, Thanks for the link. I saw some time back that someone had hacked a random number generator for cryptography applications using a sound card, but I was skeptical. What you pointed me to looks good. Too bad I'm using a Mac. Tom English

John, I couldn't possibly hold my own with you in discussion of biological evolution. You are a biologist, and I am not. But when the discussion turns to more abstract topics, I do have some things to say. "You are a hero to the 'After the Bar Closes' crowd and so apparently is Alan Fox." I've never looked at that page. The reason I'm here is that I learn a lot more engaging people with different views than I do people sharing my own view. Tom English

kairos: "First, most of the significant advancments in the genome complexity did occur in a very narrow time range;" Evidently you have the Cambrian proliferation of body plans in mind. Due to the fact that gene expression is highly nonlinear (with genetic regulatory networks, protein interaction networks, and inter-cellular communication), small changes in genotype can account for big changes in phenotype. Provided we stick to complexity in the sense of Kolmogorov (algorithmic information, as I specified above), there's no evidence to back up your claim. "second, the more complex genomes have very limited number of organisms to decently maintain RM+NS operation;" The notion that larger populations search a fitness landscape better than smaller populations is a common misconception. Under various circumstances, a small population is more likely to escape a local maximum than is a large population. "third, and very important, arguing about the yearly rate in #bits is completely meaningless without any information about the actual regularity of the search space." Almost all fitness functions are irregular in the extreme. That is, they are algorithmically random, and thus they are easy to optimize. Let's say we uniformly draw a fitness function from the set of all functions from the space of genotypes to the set of all fitness values. Irrespective of the size of the space of genotypes, 916 fitness evaluations yield with 99.99% certainty a genotype that is more fit than 99% of the genotypes. "Regularity" of the fitness landscape says little or nothing about the difficulty of optimization. A needle-in-a-haystack function and a constant function differ very little in regularity. Tom English

Silence is golden. It is the last refuge for the devout Darwinian. I love it so! "A past evolution is undeniable, a present evolution undemonstrable." John A. Davison John A. Davison

Tom English:

According to Arthur M. Lesk, the Shannon entropy of base pairs (nucleotides) in the human genome is 1.63 bits. Assuming 3 billion base pairs, the length of the maximally-compressed genome is no more than 4.89 gigabits. Assuming 3.5 billion years for that much information to enter the genome, the rate of information gain is just 1.4 bits / year.

Bad calculation. The first form of life is bacterial. It shows up about 3.5 billion years ago. The conditions for life on the early earth didn't begin until roughly that time. So I'm afraid you don't have 3.5 billion years to work with. If you look into it a little more closely you might find a different approach you can take though. In the end, the numbers might still suggest the same information increase per year. But that proves nothing. We live in a quantum world. Nucleotide A, doesn't just add itself to nucleotide B, and then A and B add themselves to nucleotide C. There are minimal requirments. Continual increase over time is not a plausible mechanism for the accumulation of information. PaV

Tom English:

Could you give us some links to actual applications of probabilistic combinatorics by ID advocates?

Why dont' you look at Fred Hoyle's, "The Mathematics of Evolution", where he calculates the probability of cytochrome C coming about by chance. It's really quite easy to demonstrate that proteins have very little likelihood of forming by chance. Do you have some idea as to how they formed by chance, as in the first cell?

This is most certainly not the case. There are many applications in which algorithmic generation of pseudorandom numbers does not suffice.

What's the point here, Tom? What are you disputing? So there are now better generators than a Monte Carlo simulator. So what. How does that affect my argument?

You seem to have problems with the Second Law of Thermodynamics. That is why there must be occasional errors in reproduction.

And you seem to have problems with the repair systems that biological forms have in place. Next question please. PaV

Well come on Tom English. Surely you aren't goimg to take that lying down or are you? We will soon see won't we? "A past evolution is undeniable, a present evolution undemonstrable." John A. Davison John A. Davison

Tom English: "I predict that as, say, quantum random number generators decrease in cost and become more widely available, evolutionary computation researchers will move away from pseudorandom number generators" Anyone with a sound card and some cheap software can get high quality random numbers due to the semiconductor-based noise generated in the electronics. For example: http://random-numbers.qarchive.org/ mike1962

Tom English I love your admission that I am a "dangerous man." What a tribute! Is that why you refuse to answer my challenges, why you pretend that I do not exist, why my name is not to be mentiond at After The Bar Closes and why you go right on blindly supporting the most failed hypothesis in the history of science, oblivious to the fact that it is a proven hoax? Being "dangerous" is what science is all about and when an adversary refuses to respond to another's challenges he has forfeited his right to be a scientist. Furthermore, he has displayed with that stance that he is an intellectual coward. The "establishment," to which you so obviously belong, is a highly organized bunch of "groupthinking" ethical degenerates who have survived for one reason only. They are hamstrung by a congenital view of the world that cannot realize that which is so obvious to those that do not suffer from that malaise. They are "born that way" and there is nothing that can be done for them. Their fate, like that of every other aspect of the world, past and present, was "prescribed." That establishment survives now exactly as it has in the past by pretending that the Darwinian fairy tale has never had any critics. You are doing nothing more than continuing that tradition when you dispense with me as you just did by describing me as a "dangerous man". Neither you nor any other Darwinian has ever answered my several challenges to your mindless dogma either here at Uncommon Descent or anywhere else. I have been summarily banned at every Darwinian sronghold on the internet. I managed just one message at P.Z. Meyers' "Pharyngula," to be greeted with "your stench has preceeded you" and instant bannishment. I love it so. You are a hero to the "After the Bar Closes" crowd and so apparently is Alan Fox. What more need be said? I don't exist over there either, just as I don't at ARN and EvC and God only knows how many other ""closed union shops" too numerous to mention. I love it so. Apparently I am not nearly "dangerous" enough. I will try to do better in future. "When all think alike, no one thinks very much" Walter Lippmann "You ain't heard nothin' yet." Al Jolson "A past evolution is undeniable, a present evolution undemonstrable." John A. Davison John A. Davison

"According to Arthur M. Lesk, the Shannon entropy of base pairs (nucleotides) in the human genome is 1.63 bits. Assuming 3 billion base pairs, the length of the maximally-compressed genome is no more than 4.89 gigabits. Assuming 3.5 billion years for that much information to enter the genome, the rate of information gain is just 1.4 bits / year" Pretty non sense. First, most of the significant advancments in the genome complexity did occur in a very narrow time range; second, the more complex genomes have very limited number of organisms to decently maintain RM+NS operation; third, and very important, arguing about the yearly rate in #bits is completely meaningless without any information about the actual regularity of the search space. kairos

According to Arthur M. Lesk, the Shannon entropy of base pairs (nucleotides) in the human genome is 1.63 bits. Assuming 3 billion base pairs, the length of the maximally-compressed genome is no more than 4.89 gigabits. Assuming 3.5 billion years for that much information to enter the genome, the rate of information gain is just 1.4 bits / year. I assume that with the long repeating segments in the genome, someone has managed to compress the genome to considerably less than 4.89 gigabits. Tom English

John, You warned me some time back you were a dangerous man, and I took you at your word. ;) Tom English

Gil: "To be fair you’d need to expand on which the source code is dependent and include the linked libraries, compiler, OS, etc." To be fair, you would code the program in the lambda calculus and submit it to a universal combinator. The universal combinator is a universal computer that has been encoded in just 425 bits. Tom English

Gil, You had a lot more in mind than to give folks a lesson in "probabilistic combinatorics." Given that your "66-character combinations" are actually permutations, I don't peg you as someone who does much work in the field. Your point was to say that if discovering such a simple program by chance is so unlikely, then how could neo-Darwinian processes obtain "biochemical and information-driven functions of living systems [that] are tightly integrated and controlled by an unimaginably complex, sophisticated, fault-tolerant, self-repairing, self-replicating computer program"? But I truly am interested in combinatorics and probability and information, and what I want to know is how many bits of (algorithmic) information there are in the compressed human genome. I would divide that number by, say, 3.5 billion years to estimate the mean number of bits of information added to the genome each year. For now, what I know is that there are about 3 billion base pairs. This indicates a gain of less than one base pair per year, and considerably less than 2 bits of information per year (because the genome is compressible). To claim that evolution has picked up less than 2 bits of information per year is hardly to worship chance. My argument is sound, and the only way you have to counter it is to oppose gradualism. That is why the notion of irreducible complexity is vital to ID. Tom English

This is for Tom English and all other Darwinian zealots. Since no one pays any attention to me anyway, here as elsewhere, I will let another with more illustrious credentials speak for me, the greatest Russian biologist of his day and, in my opinion, the greatest evolutionist of all time. "The struggle for existence and natural selection ARE NOT progressive agencies, but being, on the contrary, conservative, maintain the standard." Leo Berg, Nomogenesis, page 406, my emphasis. "Hereditary variations are restricted in number and they develop in DETERMINED direction." ibid, my emphasis. "Evolution is in a great measure an unfolding of PRE-EXISTING rudiments." ibid, my emphasis. Ergo - The Prescribed Evolutionary Hypothesis. So much for chance and accordingly so much for Darwinism in any form, the biggest, most infantile and persistent hoax in the history of science. How do you like them apples Tom? "A past evolution is undeniable, a present evolution undemonstrable," John A. Davison John A. Davison

My Hello World program serves to illustrate how quickly combinatorics create huge search spaces. To be fair you'd need to expand <stdio.h> on which the source code is dependent and include the linked libraries, compiler, OS, etc. Natural selection does indeed not have a goal, but it must produce islands of function and steadily increasing functional complexity and sophistication. As Stu Harris pointed out:

Your "Hello World" program's existence is even more wildly improbable to arrive at by chance and selection than you have stated. It is written in C. This presumes some BNR definition of the C language to begin with. How did that come about by chance and selection? It assumes a compiler written in some other language that can compile and interpret the C code. How did that come about by chance and selection? The compiler runs on an complex specified operating system, which runs on complex hardware which is composed of metallurgical, mineral, and plastic complexities that... well you get the picture. Multiply the probability of a "Hello World" program by the probability of the pyramid below it and you can come close to the un-stateable improbability of the world your program is trying to greet.

The bottom line is that there is no credible evidence that RM+NS has the creative power attributed to it in biology. It is an article of faith based on presupposition. This should be admitted, but it never is, because the entire house of cards would be seen to rest on a very shaky foundation. GilDodgen

P.S. -- The "#include" of stdio.h was truncated by WordPress above, but I did count all the characters in the line. Tom English

Gil, If the only tool you have is a hammer, then every problem is going to look like a nail. I am a computer scientist. I would love to make the genome into a program. But the "genetic program" metaphor of the 1950's simply doesn't make a good scientific model. Genetic regulatory networks are modeled in various ways. Stuart Kauffman published work on one approach, Boolean networks, in 1969, so this is nothing new. But if you did want to pursue a programming model, something like a C program would be inappropriate. Genes are weakly analogous to rules in a production system (e.g., OPS5). The rules can be arranged in a network or collection of networks. Now what is interesting about this is that one of the advertised virtues of production systems is that can add, remove, and modify rules without changing other rules. In other words, there's a high degree of modularity. And you can have rules that, in their present form, never get used, and that have little impact on the performance of the system. What I'm driving at here is that all you have to do is to change the programming model from imperative programming to rule-based programming and evolution becomes much more feasible. (In fact, our mutual friend in La Jolla and I were once going to work on evolution of rules, but I decided on another approach.) Of course, the genetic programming community likes to work with LISP, a language with almost no syntax rules. I have to say that in the example you link to you repeat a fallacious argument that creationists advanced a number of years before ID came on the scene. You cannot argue against neo-Darwinism by making it into something it is not. Neo-Darwinian evolution has no goal. Yet you set up a goal and argue that it would be impossibly hard to achieve. If you want actually to challenge neo-Darwinian theory with a counting argument, you have to take it as given by scientists. Another point about your example is that, even if I count 7 bits for each and every character, I get only 525 bits. Thus if an evolutionary process could pick up only a bit of information per year in interaction with the "environment," it would have the C program in a mere 525 years. You get a "deafening cry of design" only if you slip in an assumption that gradual gains in information are impossible. I should mention that a fair amount of your code is not necessary. Here is a 280-bit program that outputs something: #include main(){printf("H");} This is an approximation to your full C program, and 2 ^ -280 is well below Bill Dembski's universal probability bound. Furthermore, in a 64-character LISP implementation, (print "hello, world!") is a 144-bit program. A question I have about your C program is, if I came up to you and said I generated it all at once by chance, how would you compute the complex specified information to show that I had not? Tom English

Just as a reflection, it seems to me that it would be a great thing if biology students were required to take a class on computer programming.... Anyone having gone through that experience would instinctively “know” that RM+NS just isn’t going to get it done.

Speaking as someone with a bachelors in Computer Science (software engineering concentration), with two decades of programming experience, and as someone who completed the undergraduate work in pre-med, the evolutionary mechanism of RM+NS makes very good sense to me. You have to understand that genes are not like sections of computer code. I think you are going to underestimate the usefulness and flexibility of RM+NS if that's your analogy. I think that's why engineers and programmers get hung up on RM+NS: they are thinking about evolution through their experience of programming languages and engineering tasks. It's a natural human trait to think about new information through the lense of past experience, but in this case, I think the analogy doesn't hold very well and it leads to some misleading conclusions. BC

Sorry Tom English, but it won't wash. There was never a role for randomness in any aspect of phylogeny just as there is now no role for it in ontogeny. To claim that which cannot be demonstrated is intellectally dishonest and pure fantasy. "Neither in the one nor in the other is there room for chance," Leo Berg, Nomogenesis or Evolution Determined by Law, page 134 Please note the complete title of Berg's opus magnus, in my opinion the most significant single work in all of the evolutionary literature. We will soon be discussing Bergian evolution, as the Darwinian myth descends into oblivion to join the Ether of Physics and the Phlogiston of Chemistry, nothing more than a footnote. How it has survived as long as it has is a mystery. It should have died twelve years after its inception with the publication of St George Jackson Mivart's book with the tongue-in-cheek title - Genesis of Species. That is roughly when the Ether of Physics bit the dust and a century after Phlogiston had also gone the way of all inventions of the human imagination. It is hard to believe isn't it? I love it so! Who is next? "A past evolution is undeniable, a present evolution undemonstrable." John A. Davison John A. Davison

PaV: "And as to 'randomness', anyone doing any kind of programming at all knows that the only kind of 'randomness' you want in a program is the kind the 'program in', for example, a Monte Carlo simulation program." This is most certainly not the case. There are many applications in which algorithmic generation of pseudorandom numbers does not suffice. Pseudorandom number generators implement small algorithms, and this means that they can supply only a very limited amount of algorithmic information. There are web sites at which you can get random numbers obtained through measurement of natural phenomena, e.g., http://www.fourmilab.ch/hotbits/ http://www.random.org/nform.html I have shown in published work that, in practice, only a small fraction of all possible searches can be implemented with pseudorandom number generation. I predict that as, say, quantum random number generators decrease in cost and become more widely available, evolutionary computation researchers will move away from pseudorandom number generators (though they will still use pseudorandom generators to achieve repeatable results while developing software). PaV: "Thus, it is not surprising that biological forms use 'randomness' here and there; but that 'randomness' is very likely 'programmed in'. So, paraphrasing the world-renowned Richard Dawkins, 'Biology is the study of life forms that give the appearance of randomness.' I believe that this strikes much more closely to the truth of life." You seem to have problems with the Second Law of Thermodynamics. That is why there must be occasional errors in reproduction. Nothing in nature is "trying" to introduce random errors. Nothing in nature is faking them. Physics necessitates random errors, and they are a source of information in evolution. I think the biggest problem IDists have with this is that they don't know enough information theory to understand that an entity is most informative when it is "most random." Tom English

DLH: "Here are some interesting searches on Probalistic Combinatorics. Look forward to combinatorial specialists adding serious reviews." I know enough to know what is and what is not probabilistic combinatorics. (I have published a bit on the properties of decision trees corresponding to randomized search algorithms.) Gil has done only elementary counting -- there's no "probabilistic" to it. I would like to see some links to actual applications of probabilistic combinatorics by ID advocates. Tom English

DLH: "Here are some interesting searches on Probalistic Combinatorics. Look forward to combinatorial specialists adding serious reviews." Well, probabilistic combinatorics is a fancy term. And it's also an interesting field, for those of us who have actually looked into it. (I have published a bit on the properties of decision trees corresponding to randomized search algorithms.) The problem is that Gil does only elementary counting -- there's no "probabilistic" to it. Could you give us some links to actual applications of probabilistic combinatorics by ID advocates? Tom English

But any self evidence is almost never a primary thought; it is the result of a reasoned judgement on which among several explications for a given event is really correct. Take a similar example of design recognition: a person sees an alien articraft and is able to recognize it as a non-natural but designed thing, and this without any error (you would see "self evident"). But how could have he produced this judgement without fast and unconscious rejecting the natural option and then accepting the design one? kairos

kairos I never looked at it that way I guess. You aren't the first to be thorougly confused by my statements. Life is like that. "A past evolution is undeniable, a present evolution undemonstrable." John A. Davison John A. Davison

#7,11 "Design is not an inference and should never have been presented as such. "I have always regarded Intelligent Design as self-evident which is all that I meant or intended". John, I'm a bit confused by your statements. If one finds that ID is self evident this is just because he has actually done an inference to the best explanation of the biological world. kairos

Carlos I have always regarded Intelligent Design as self-evident which is all that I meant or intended. Each of us has his own interpretation of the world he peruses. That is hardly a "bone of contention." I believe it is known as freedom of speech, and is covered by the United States Constitution. Correct me if I am wrong. "A past evolution is undeniable, a present evolution undemonstrable." John A. Davison John A. Davison

Design is not an inference and should never have been presented as such. I'll leave this matter between you and Dembski. If I understand your respective views correctly, this is a bone of contention. Carlos

ID exam follow-up question: Had I written, "A microbe did not mysteriously mutate into Mozart and his music, and most people, mercifully, are much too smart to swallow this a silly idea," would you have detected alliterative design? How many alliterative M's would it take to make design an obvious, slam-dunk conclusion? Explain. GilDodgen

ID exam question: Is the "m" alliteration in the last sentence of my post by design or by chance? Did you detect it? Explain. GilDodgen

Design is not an inference and should never have been presented as such. It is a mandatory and self-evident feature of every living thing, its development and its evolution. A role for chance has been completetly eliminated from consideration as being involved in either ontogeny or phylogeny. If not chance then what? I say the Prescribed Evolutionary Hypothesis is the only conceivable alternative to the most failed hypothesis in the history of science. I regard it as firmly established and will until it has been discredited by controlled laboratory experiment. So far everything we are learning pleads for the great antiquity of biological information, much of it until recently thought to be of recent origin. The source of that information has always been the chromosome which is the seat of all meaningful information. How much of that original information still remains is problematical but that it once existed cannot be denied except by those that worship the Great God Chance. Variations in the particulate gene had nothing to do with either ontogeny or phlogeny which is why all living thimgs are so remarkably similar at that genetic level. Evolution, a process no longer in progress, was largely, if not entirely, a "position effect" phenomenon utilizing a common mechanism for which a controlled modification of the intrachromosmal environment served to alter the rate and timing of key developmental events. All that allelic (Mendelian) mutations ever did was introduce pathological modifications of an otherwise predetermined sequence. "The theory of the genes and of the accumulation of micromutants by selection has to be ruled out of this picture." Richard B. Goldschmidt, The Material Basis of Evolution, page 396 "A past evolution is undeniable, a present evolution undemonstrable." John A. Davison John A. Davison

Here are some interesting searches on Probalistic Combinatorics. Look forward to combinatorial specialists adding serious reviews. DLH

Re: (4) Strict empiricism may assist the criticism of NDE, but it won't assist the promotion of any alternative. On the contrary, strict empiricism turns science into stamp collecting. ID, after all, is not strictly empirical. One does not observe design; one infers it. (That's why one of Dembski's books is The Design Inference, right?) On the other hand, if the demand for empiricism is relaxed, then one can consider different possible inferences, evaluate the probability of each one, the relationship between different lines of evidence, and the explanatory potential of different theories, etc. Carlos

Natural selection is the cornerstone of Darwinism. It has been since its inception. It was a logical argument devoid of empirical support at the outset and still is. Advocates of NS point to sequence homologies as evidence of common descent and then argue that NS would preserve advantagous genetic changes. But what observations actually document multiple mutations leading to novel complexity? The out is that such occurences take millions of years so it is not realistic to expect to see such changes in a lab envirnment even for rapidly reproducing unicellular organisms. It it a good means of isolating the theory from testing. What we are left with is arguments based on homologous genes. pk4_paul

I am about 2/3 through Sean Carroll's book, Endless Forms Most Beautiful. In it he describes an incredible array of what he calls switches that turn on and off the various genes to produce each cell in our body during gestation. It sounds like there are tens of thousands or more of these switches in the genome all orchestrated in a perfect way to form the embryo. He says it would take a book of several thousand pages to describe the sequences of switches for the human body, each page densely packed with information. Yet he ascribes all this to be a product of chance. I think it is obvious from the comments we often get here that many people are so emotionally tied up in a world view that they will believe anything to support this worldview. This is odd when what is being discussed is science but what they choose to believe is driven by their worldview. The combinatorial processes that could produce an embryo are staggering and to believe that it could just happen by chance is really an example of their worship of their god of chance. In their eyes their god is more powerful than the God of traditional religion. By the way Carroll's book is an exercise in what I call the Kindergarten/Graduate School approach. The author starts out trying to explain the material as to an average person using easy to understand ideas and then get so complicated that it would take a graduate degree to follow what is said. It is an interesting book but in no way is there anything in it that supports Darwinism other than the author's personal comments. Insects and humans have identical developmental genes so Carroll says there must be a common ancestor that existed prior to the Cambrian Explosion that had these genes. But as of today nothing has been found prior to the Cambrian Explosion that could be this universal predecessor that gave rise to all the phyla of the Cambrian. jerry

You mention that you're a computer software engineer. Here and there on this board, it's been commented on that 'engineers' seem to be naturally inclined to the ID perspective. The other element here is, of course, computer programming. Just as a reflection, it seems to me that it would be a great thing if biology students were required to take a class on computer programming. I did, back in the 70's; the good, old days when you had to punch your own cards, wait for some big main-frame computer, and hope to high heaven that you had all the cards in the right order. Anyone having gone through that experience would instinctively "know" that RM+NS just isn't going to get it done. And as to "randomness", anyone doing any kind of programming at all knows that the only kind of "randomness" you want in a program is the kind the "program in", for example, a Monte Carlo simulation program. Thus, it is not surprising that biological forms use "randomness" here and there; but that "randomness" is very likely "programmed in". So, paraphrasing the world-renowned Richard Dawkins, "Biology is the study of life forms that give the appearance of randomness." I believe that this strikes much more closely to the truth of life. PaV

Gil, sometimes we stand back and see the "breathtaking inanity" as it seems to us, of conssidering RM+NS the origin of the biosphere. Others stand back and marvel at the "breathtaking inanity" of ID. I wonder what distorts at least one of the two pictures? idnet.com.au