Uncommon Descent Serving The Intelligent Design Community

Why Mathematicians, Computer Scientists, and Engineers Tend to be More Skeptical of Darwinian Claims


Larry Moran’s presentation in a comment in Granville Sewell’s UD post, I found not particularly persuasive, for the following reasons. I’m not interested in definitions of science; I’m interested in how stuff actually works. I’m perfectly amenable to being convinced that the complexity, information content, and machinery of living systems can be explained by stochastic processes filtered by natural selection, and I would not even demand hard evidence, just some rigorous argumentation based on the following:

1) A particular aspect of any living system that displays a machine-like function (such as a ribosome).
2) Some specifics about what random genetic changes (of any type) would be required to engineer intermediate forms.
3) A reasonable estimate about the likelihood of these random changes occurring.
4) Another reasonable estimate about the likelihood of the hypothetical intermediate forms providing a statistically significant survival value.
5) Some kind of evidence or even reasonable conjecture that the number of individuals and reproductive events could provide the requisite probabilistic resources. Appeals to “deep time” are irrelevant.

These are the kinds of challenges that those of us involved in mathematics, computer science, and engineering tend to present, and the kinds of questions we tend to ask, because we must demonstrate that our stuff can actually work in the real world, or at least that it has a reasonable prospect of working in the real world. That’s why many of us tend to be skeptics.

DLH makes an excellent point which has a name: entropy. Engineers have to deal with it in all their designs and even with our best, most carefully contrived efforts we can't stop it. All we can do is slow it down. In the matter of life I totally understand extinctions. That's fully explained by entropy. What I don't understand is how some cell lines manage to slow it down enough that they've persisted for billions of years. In computer design we've developed highly protected core software that reset what evolution hath wrought back to a known good state. Typically this is implemented by a read-only memory chip that instructs the hardware to load a protected image of the factory-original software configuration. The factory original is, because of its size, usually supplied as a protected portion of a hard disk drive and/or on removable media such as a DVD disk. To insure even greater chance of catastrophe recovery we store redundant copies in different locations such that if a fire destroys one copy others will still be available. Since human design more often than not is eventually found to be an analog of something in the design of life my best guess as to how some cell lines manage to slow down entropy enough to survive for millions or even billions of years is by employing protected core code and triggers which serve to restore the organism to a known working state. This is an important question. I'll write a separate article today to address it. DaveScot
mathematicians are naturally skeptical of darwinism because mathematicians are inherently risk-averse when it comes to making statements without rigorous proof. the level of rigour in some evolutionary psychology speculations is apalling by mathematical standards. there is much better evidence for numerous ideas in mathematics which mathematicians continue to call conjectures due to absence of formal proof, then there is for some of the evolutionary psychology speculations. tfoo
one brow said,
My take on the original post: Mathemeticians, engineers, and computer scientists all work in fields dominated by the type of thinking that goes into a formal system. Mathematics is a full-fledged formal system, while the other two are highly rule-driven with an objective to create. They are essentially fields where you use deductive thinking, working from the top down.
Yes, I think that one of the reasons why many of us went into those fields (I am a mechanical engineer) was that the laws and fixed rules in those fields gave us feelings of security, order, certainty, and predictability. That is not to say that those fields are not challenging -- it is often a challenge to figure out ways to apply the laws and fixed rules. Larry Fafarman
My take on the original post: Mathemeticians, engineers, and computer scientists all work in fields dominated by the type of thinking that goes into a formal system. Mathematics is a full-fledged formal system, while the other two are highly rule-driven with an objective to create. They are essentially fields where you use deductive thinking, working from the top down. Science is the process of building from the bottom up, using inductive thinking. The prupose is not so much to create but to uncover. It's very natural that the people who were attracted to fields that relied on deductive thinking in the first place are more likely to see the top-down version of reality, with an outside force providing guidance and rules, as a more acceptable vision of the world. I’m perfectly amenable to being convinced that the complexity, information content, and machinery of living systems can not be explained by stochastic processes filtered by natural selection, and I would not even demand hard evidence, just some rigorous argumentation based on the following: 1) An analysis of the unlikelness of evolution that accounts for all 10+ known processes by which the genome accumulates variances, the 3 methods by which that variance is selected, and demonstrates true independence for each facrot before multiplying the factors' probabilities together. 2) A reason that scaffolding and similar methods would not be able to create irreducilbe complexity naturally. 3) A way to measure CSI that is internal to the information contained in the string itself. For example, it should be able to determine with 95% accuracy whether a given four-letter code in a string of 3000 characters either comes from a coding section of DNA or was randomly generated. 4) A precise limitation on the features that can or can not appear in a living organism over a trillion generations. 5) A list of the precise interventions that were made in the history and the time frame in which they were made. Appeals to “some time in the past” are irrelevant. one brow
Atom - thanks for the heads-up on the other post (this one is too far away! :-)). Once an allele becomes fixed, yes there is no fitness benefit so the fitness becomes 1. And yes, this makes perfect sense. My somment about the fitness being re-set was wrong. The situation Haldane imagines is where a change in the environment reduced juvenile survival, and this is restored though selection. His argument boils down to saying that if the reduction is large enough, the growth rate goes below 1, so the population will decline, and may go extinct. Nunney's model is almost exactly the same in this regard, except that he has a continuously changing environment (it's also different in that it has better ecological dynamics, but that's not critical for the general point). Bob Bob O'H
Bob O'H, still there? Atom
I cited Nunney to clarify the point that Haldane’s dilemma only works in a deteriorating environment. It’s something that isn’t obvious from Haldane’s paper. If most environments are stable (at least with regards to the fitness surface), the dilemma goes away because beneficial mutations increase fitness - in Haldanes model (IIRC) the fitness gets re-set when an allele becomes fixed. That’s an odd feature of the model, but is best interpreted as being equivalent to a decline in the environment.
Bob, Let me know if I misunderstand your point. Are you saying that in Haldane's model once an allele reaches fixation its fitness becomes 0 (ceases to confer a fitness benefit) and that Nunney's model disagrees with this assumption? If not, then disregard the rest of this and please clarify. If so, I think Haldane's assumption makes perfect sense, since a trait increases fitness over that of your competitors (other organisms in your population.) If everyone has the same allele (fixation), then there is no more fitness advantage. Please correct me if I misunderstood you. Atom
DaveScot said,
I think you’re quite right that this is a major factor in why prehistoric evolutionary biology is so vociferously defended by those with a vested interest in it or similarly useless intellectual pursuits.
I wouldn't call paleontology -- which to some people is synonymous or nearly synonymous with "prehistoric evolutionary biology" -- a "useless intellectual pursuit." Knowledge is good for its own sake. However, the Darwinists have been claiming that the US is going to lose its international technological competitiveness if Darwinism is not taught dogmatically in schools. Paul R. Gross, co-author -- with Barbara Forrest -- of "Inside Creationism's Trojan Horse" and lead author of the Fordham Institute (no connection to Fordham U.) report on state science standards, threatened to drop Ohio's overall science grade from a B to an F because of Ohio's evolution lesson plan, even though evolution counts for only 3 points out of 69 in the Fordham grading system -- see http://im-from-missouri.blogspot.com/2007/12/fordham-reports-lead-author-is-co.html Larry Fafarman
Larry Good point. Science and engineering that produces practical benefits easily seen by laypersons is not really subject to threat when allocation of finite public funding is prioritized and distributed as compared to science and engineering that does not produce easily discernable practical benefit. One program that comes immediately to mind in this regard is the Superconducting Super Collider. Congress critters couldn't see the cost/benefit of the big atom smasher as superior to that of the International Space Station, couldn't afford to fund both, so the SSC lost out. Historic biology is of far less practical import than high energy physics. The public would refuse to grant much funding to historic biology if they realized how useless it is. I think you're quite right that this is a major factor in why prehistoric evolutionary biology is so vociferously defended by those with a vested interest in it or similarly useless intellectual pursuits. DaveScot
The Machine v. 1.1* Two people are looking at a finely-tuned, highly sophistocated machine that produces massive amounts of complex, functioning items. 1: “It’s remarkable what purposelessness can do.” 2: “You think this machine came about and operates without purpose?” 1: “Yup.” 2: “That’s incredible. What makes you think that?” 1: “Come on. Everyone knows that everything in life is purposeless.” 2: “Why do you say that? All machines I know of have been designed and built for a purpose, by intelligent beings.” 1: “Yeah, but the intelligent beings came about through a process that’s purposeless.” 2: “How do you know?” 1: “Come on. Everyone knows that.” 2: “Let me take another approach. Can you show me a working model of a purposeless machine that generates any type of functioning thing?” 1: “Here. Here they are." [Pointing to Avida, etc.] 2: “But these all have active information and an overarching purpose built-in! How can you claim that these are models for this machine?!” 1: “Well, for a living, I study the output and the few visible portions of this machine. All the books and all the important people in the field say that the machine is entirely purposeless. So there. Who are you to question that?” 2: “I design and build machines every day. When we have an idea for a machine, before committing large amounts of money to its development for production, we always need to prove that it will work by designing and building a working model. You want me to commit to the idea that this thing operates completely purposelessly, yet you haven’t shown a proof-of-concept working model for your idea of how it works. Sorry, no can do.” * With a few minor but clarifying fixes in the wording of my little story at comment 5 above. j
Also, I think that scientists in non-biological sciences -- e.g., inorganic chemistry, physics (Lord Rutherford's "stamp collecting" statement notwithstanding), geology, and astronomy, and even the social sciences, e.g., anthropology, sociology, and psychology -- may feel that by defending Darwinism they are defending science in general. Pure scientific research often has no immediate economic benefit and scientists may feel that research funding may be reduced if the attacks on Darwinism cause the public to lose faith in science. On the other hand, the work of engineers and so-called computer "scientists" often has immediate economic benefit and so they may not feel threatened in this way by attacks on Darwinism. Larry Fafarman
In addition to asking why mathematicians, computer scientists, and engineers tend to be more skeptical of Darwinian claims, IMO we should also be asking why biologists tend to be less skeptical of Darwinian claims. The lower skepticism of the biologists makes the skepticism of the mathematicians, computer scientists, and engineers seem great by comparison. Biologists have an inferiority complex because of the kind of attitude expressed by Lord Rutherford: "All science is either physics or stamp-collecting." Because of this inferiority complex, biologists have been boasting with a religious-like zeal that biology has something that other branches of science do not have, a grand central overarching supreme unifying "theory of everything," Darwinism. Larry Fafarman
DLH - thanks for the tip on the e, but I can't see one. :-( Perhaps it'll turn up on this message. I cited Nunney to clarify the point that Haldane's dilemma only works in a deteriorating environment. It's something that isn't obvious from Haldane's paper. If most environments are stable (at least with regards to the fitness surface), the dilemma goes away because beneficial mutations increase fitness - in Haldanes model (IIRC) the fitness gets re-set when an allele becomes fixed. That's an odd feature of the model, but is best interpreted as being equivalent to a decline in the environment. I'm well aware of Remine's complaints about Nunney. Whilst I think Nunney could have been politer, I don't think Remine has a case. The model is adequately described in the paper, so Remine could code it himself. It should be fairly easy in a high-level maths language. Bob Bob O'H
Semiotic 007 at 53 Your argument fails to consider the lemming effect in science and paradigm shifts after one brilliant person shows the way. e.g., the Aristotelian academicians were collectively opposed to Galileo and did all they could to stop him. Yet Galileo showed the evidence and Aristotle no longer reigns. See Mathematics of Evolution at ResearchIntelligentDesign.org Sir Fred Hoyle: spent a major part of his life digging into the mathematics of evolution. He observed:
If one proceeds directly and straightforwardly in this matter, without being deflected by a fear of incurring the wrath of scientific opinion, one arrives at the conclusion that biomaterials with their amazing measure or order must be the outcome of intelligent design. No other possibility I have been able to think of...[7]
"Published in his 1982/1984 books Evolution from Space (co-authored with Chandra Wickramasinghe), Hoyle calculated that the chance of obtaining the required set of enzymes for even the simplest living cell was one in 10 ^ 40,000. Since the number of atoms in the known universe is infinitesimally tiny by comparison (10 ^ 80), he argued that even a whole universe full of primordial soup would grant little chance to evolutionary processes." Hoyle further refined his observations in Mathematics of Evolution, building on published population genetics works. Yet who today dares to take Hoyle's mathematics and observations in the face of the wrath of evolutionists in power? See especially: Fred Hoyle, Mathematics of Evolution, (1987) University College Cardiff Press, (1999) Acorn Enterprises LLC., ISBN 0-9669934-0-3 Hubert P. Yockey, Information Theory, Evolution, and the Origin of Life 2005 Cambridge University Press ISBN 13 978-0-521-80293-2 John C. Sanford Genetic Entropy & the Mystery of the Genome. (2005) Ivan Press. ISBN 1599190028. DLH
Bob O'H at 51 Thanks for the reference to: Nunney, L. 2003: The cost of natural selection revisited. — Ann. Zool. Fennici 40: 185–194. (free screen quality pdf) I understand the "cost" in generations per fixed mutation is dominated by the beneficial mutation rate M. In "Genetic Entropy and the Mystery of the Genome", John Sanford summarizes reported beneficial to harmful mutation ratios of 1 to 10,000 or 1 to 1 million. It appears Nunney was using very high (optimistic) beneficial mutation rates. I don't see how his study makes any serious change to the strong limitations of Haldane's Dilemma. Nunney may relax it by one or two orders of magnitude - IF the high beneficial mutation rate is justified. The 150 million changes between human and chimp is still many orders of magnitude greater than the maximum available time. PS Walter Remine reports:
Evolutionary geneticist, Leonard Nunney, reported on his computer simulation that evolves dramatically faster than the Haldane limit. Walter ReMine contacted Nunney and requested a copy of Nunney's software for detailed study of its results and methods. Nunney declined, saying he would not share his software with "people who do not publish in peer-reviewed journals" -- which is evolutionist-code for 'anti-evolutionists'. Given evolutionist reluctance to have their results verified, Nunney's simulation must be viewed dimly, especially since it contradicts other simulations.
Bob O'H et al. To correct a post see if you can click the "e" underneath your name and date on the left of the post. DLH
well if that's the case you can end up with bloat but you won't get evolution. In real life, selective pressures are low, the possibility of a mutation that adds functionality is very low, the extra functionality would be very small and it still has to justify the cost of maintaining that extra complexity -it is small but not when compared to the extra benefit. On the other hand, a mutation that removes complexity (without hurting it) could be a big energy win and natural selection would favor that. That's why it is hard for me to see how evolution could work. ari-freedom
You are ignoring the fact that recent research has found that much or most of the genome is indeed functional and that there is much less “junk DNA”
I made no claim whatsoever about functionality in the genomes of biota, magnan. And you have not addressed the fact that most genetic programming systems rely heavily on insertion of random code. Casual observations about probability of success have little value when there are empirical data to go on. Semiotic 007
Ari, I hate engaging in seat-of-the-pants "reasoning" about topics I know little about. Do you know how much of the resource requirement for cellular reproduction is accounted for by DNA replication? My guess is that it is little. It's not obvious to me that there is significant selective pressure for a small genome. Replication errors, on the other hand, are a thermodynamic necessity. Semiotic 007
Matteo, In my opinion, no one is expert in design inference. There has certainly been some work in the past 10-15 years (a shorter period than Darwin spent coming by his theory of evolution) to develop a theory of design inference, but there is not yet a stable body of knowledge. Dr. Dembski's last paper on design detection through measurement of complex specified information is a far cry from what he presented in The Design Inference. His recent collaborative work with Bob Marks does not mention CSI, but is developed in terms of endogenous, exogenous, and active information. While an IDist would see active information as analogous to CSI, its definition is radically different. It appears that Dr. Dembski has left CSI behind. Recall that Prof. Behe indicated in Darwin's Black Box that his "evolutionary" definition of irreducible complexity was stronger than the original. That is quite significant, because irreducible complexity is a matter of degree in the revised definition, and is all-or-nothing in the original. Dr. Dembski has since written a long paper offering his own revised definition, and it differs substantially from Behe's. ID theory has shifted radically and rapidly in its formulation of design inference. I cannot see that you have a basis for claiming that engineers, computer scientists, and mathematicians have expertise in design inference that other people do not. Again, there are investigators, but there are as yet no experts. Best wishes and good luck to Dembski, Behe, and others in their research. Semiotic 007
Semiotic 007 (#55): "Hmm. So C-language code inserted after return and break statements is problematic? How often do you suppose a randomly-generated condition for an if statement would be satisfied in practice? There are more ways to introduce junk code into programs than are apparent to the casual observer." So how much of the code is functional as opposed to nonfunctional? What if the injected error creates a jump statement into a nonfunctional area after return or break statements (ignoring issues of the particular language)? The fact remains that there is a high likelihood that a random change inserted into a random point in a software program will cause a problem and degrade the software machine (if not cause catastrophic failure and stop it or send it into an endless loop). "Of course, IDists don’t care much for the claim that much of the genetic code of biota is nonfunctional. But it is commonly the case that much of the code in genetic programs is nonfunctional." You are ignoring the fact that recent research has found that much or most of the genome is indeed functional and that there is much less "junk DNA" than previously supposed by Darwinists. It is turning out to be an incredibly complex interwoven network with many levels of coding. This has been discussed in numerous other threads on UD, and is directly contradictory to the predictions of Darwinists. magnan
Semiotic, I used computer code simply for discussion of my microsoft example. From what I understand, gene regulation networks function more like electrical engineering logic circuits. But this is irrelevant. Again, we're not talking about specific instantiations of machines, but rather information processing and information flow principles in general. These apply to Turing Machines, logic networks, carbon-based machinery, wood-based machinery, "silly putty"-based machinery, or any other type of machinery you'd like to create within this universe. A simple hammer comes in handy when you are dealing with different types of the general "nail" class. (Or perhaps you'd like to argue that carbon based machinery is somehow special and exempt from the general laws?) Atom
following what I said in 59. Any mutation that increases complexity (I'm not using the term CSI here), even if functional has an energy cost to support the extra complexity. This would make evolution very hard to pull off. Natural selection will only favor mutations that simplify the system. Now in a virus, you don't have this problem so much because it just leaches off all the hard work of a living organism. But a living thing needs to have the 'capital' to make the 'investment.' ari-freedom
Atom, "If the only tool you have is a hammer, then every problem's going to look like a nail." A senior professor in engineering pointed that out to me many years ago, back when I was a green assistant professor, and I took it to heart. Many people take the "genetic program" metaphor literally. Computing professionals seem particularly susceptible to the error. It is revealing that they usually speak in terms of programs for general-purpose computers, which they feel they know and understand, rather than for robots with custom hardware and firmware. Certainly the analogy to a program for a robot is much stronger than the analogy to a program for a general-purpose computer, yet it eludes most who want to speak in terms of programs. The "genetic program" does not come close to controlling the state of the "machine" as the program of an electronic digital computer does. And it appears that in morphogenesis of multicellular organisms a great deal of the information comes from the environment, rather than from the genome. Metaphors can be useful for those who remember they are metaphors and who understand their limitations, but can be highly misleading for those who take them literally. Semiotic 007
@Matteo 60: Well put. Atom
Who was Darwin? All he had was a theology degree. He made evolution into a new religion. Biology is not a science because of this but that will change once Darwin is taken out and engineers are put back in. ari-freedom
Semiotic 007, I'm going to reword your post a bit to demonstrate that while the structure of it is well thought out, the assumptions behind it merely beg the question: "The majority of degreed biologists have completed no college course work in the engineering sciences. Virtually all have college physics under their belts. Some studied chemistry in college. Relatively few enrolled in college courses in engineering or systems design. Among “expert” critics of scholarly fields not their own, at most one in a thousand makes a substantive contribution. If the biological sciences should happen to be chock-full of biologists who have all caught engineers, computer scientists, and mathematicians in fundamental error, then it would constitute a singular event in the history of science. Most of us who have actually contributed to the bodies of knowledge in our own fields know that the most reliable sources of information in other fields are the people devoted to investigation of those fields. Scientists are sometimes in error, but as consumers of scientific knowledge we have no choice but to play the odds. The odds of getting useful explanations of the design of life forms on earth are much greater when one goes to people who have studied the matter intensively and have worked their way to consensus than when one goes to people who have little or no higher education on the matter. Maverick geniuses are very rare, and if you ever seem to be surrounded by them, you can count on it that appearances are deceiving." Bottom line, the assumptions you make are entirely upside down. It is the biologists who have no special expertise when it comes to evaluating design inferences. They have no special standing to tell those who do have expertise in fields directly impinging on design, probability, and information theory to simply sit down, shut up, and listen to what the non-experts have to say. IMO, the non-expertise of evolutionary biologists in any of these crucially pertinent fields becomes more and more annoyingly manifest with each passing year. Matteo
semiotic: bloat has a big cost. It takes up energy and resources to do nothing. Maybe not a problem in the digital world but in the real world I would expect natural selection to eliminate the bloat. Why didn't it? Either the code really is functional and the organism is more more complex OR natural selection doesn't work very well. ari-freedom
All scientific beliefs are tenuous. Anyone who does not want to play the odds in belief formation should go somewhere other than science. I personally recommend religion and philosophy. Semiotic 007
And BTW, I have done university coursework in the life sciences. Atom
1 2

Leave a Reply