|
Author
|
Topic: Sources of Randomness in Natural Selection
|
William A. Dembski
Member
Member # 7
|
posted 10. January 2003 10:31
The Sources of Randomness in Natural Selection:
Who's Whispering in Natural Selection's Ear?
By William A. Dembski
About a year ago I was interviewed with Eugenie Scott on a PBS program titled Uncommon Knowledge ("Darwin Under the Microscope"). The host, Peter Robinson, asked how a random evolutionary process could bring about the complexity we see in biology, and then brought up the famous chestnut about monkeys typing Shakespeare. Eugenie's answer was that evolution isn't like randomly typing monkeys acting all alone but rather like randomly typing monkeys with someone standing at the monkey's shoulder erasing any mistakes that the monkey makes.
Now there is an obvious problem here -- how does who- or whatever is at the monkey's shoulder know what to erase? If the monkey is supposed to be typing the works of Shakespeare, then the monkey's mistake-correcting helper presumably must know the works of Shakespeare already, which overthrows Eugenie's point, which was to get the works of Shakespeare without Shakespeare or any other designing intelligence.
Turn now to natural selection as it operates in biology. According to Richard Dawkins evolution should not be conceived as fundamentally random. Yes, there's random variation, but, as he puts it in the Blind Watchmaker, natural selection is "quintessentially nonrandom." Dawkins is therefore saying that certain self-replicating systems -- simply in virtue of the way nature is constituted and the natural laws that govern it -- will be better at surviving and reproducing than others.
This may remove the arbitrariness and randomness in the day-to-day operation of natural selection but not in the structure of nature that determines what nature can select. Return to Eugenie's mistake-correcting helper. If the helper is not a designing intelligence, then that helper is him/her/itself going to need some help in knowing what to erase. There is a regress here, and it confronts natural selection operating in biology as well.
Why should natural selection choose one organism to survive and reproduce and not another? Even here randomness cannot be wholly avoided because which organism gets to survive can simply be a matter of chance (the "better fit" organism may lose out simply by slipping on a banana peel and thereby breaking its neck). But the deeper point is that even if on average the "more fit" survive, what determines which organisms are "more fit"? How was nature structured to facilitate a complexity-increasing form of evolution driven by natural selection? Apart from design, sources of randomness will inform natural selection. But in that case natural selection itself becomes essentially (though perhaps not quintessentially) random.
IP: Logged
|
|
Evan
Member
Member # 164
|
posted 10. January 2003 17:01
I got back from teaching this afternoon and intended to reply to the ideas brought up by the monkey analogy in this thread. However it seems like the 2nd post here (which I wanted to refer to) is now gone. I thought there were some good points there, and I wanted to respond. What happened? Can we get that post back?
IP: Logged
|
|
Moderator
Administrator
Member # 1
|
posted 10. January 2003 17:45
Evan,
Frances post was removed because it was posted only 1.5 hours after Dembski posted his initial message. In addition, Frances has been warned about being the first poster to respond to a new thread. Further, he has been exceeding his daily posting limits.
Frances has been warned about these things in the past and action needed to be taken. Sorry that it had to be on a post that was useful to you.
He is free to post his message again tomorrow.
IP: Logged
|
|
andyg
Member
Member # 415
|
posted 10. January 2003 18:16
quote:
Turn now to natural selection as it operates in biology. According to Richard Dawkins evolution should not be conceived as fundamentally random. Yes, there's random variation, but, as he puts it in the Blind Watchmaker, natural selection is "quintessentially nonrandom." Dawkins is therefore saying that certain self-replicating systems -- simply in virtue of the way nature is constituted and the natural laws that govern it -- will be better at surviving and reproducing than others.
This may remove the arbitrariness and randomness in the day-to-day operation of natural selection but not in the structure of nature that determines what nature can select. Return to Eugenie's mistake-correcting helper. If the helper is not a designing intelligence, then that helper is him/her/itself going to need some help in knowing what to erase. There is a regress here, and it confronts natural selection operating in biology as well.
I think Dembski is taking Scott's analogy a little too literally. Selection is the cumulative process of replicators replicating in a particular environment. What happens depends on the nature of the replicator and the nature of the environment. If you grow of population of bacteria in antibiotic, and only 1% of the bacteria are resistant to that antibiotic, then over time, that sub-population of bacteria will become betetr represented in the population. When Dembski asks "how does who- or whatever is at the monkey's shoulder know what to erase?" he is completely missing the point. The "selecting force" for want of a better word is unique to the combination of the antibiotic containing medium and the bacterial population. If you used a different antibiotic and a different population of bacteria, you would get a different outcome.
quote:
Why should natural selection choose one organism to survive and reproduce and not another? Even here randomness cannot be wholly avoided because which organism gets to survive can simply be a matter of chance (the "better fit" organism may lose out simply by slipping on a banana peel and thereby breaking its neck).
This is a trivial point. Selection is best viewed by looking at large populations over generations, not over very few numbers of individuals over few generations. That's why people get hung up on the apparent "tautology" of the survival of the fitetst. The fittest in a given environment do not survive axiomatically. They simply have the best chance of surviving. Pennock has pointed out this fallacy with respect to the phrase "may the best team win".
quote:
But the deeper point is that even if on average the "more fit" survive, what determines which organisms are "more fit"?
Again, the interaction between the genotypes in a population and the environment determines what is "fit" in a particular instance. In an ampicillin-containing medium, ampicillin-resistant bacteria will be more "fit" than kanamycin-resistant bacteria. The opposite will be true in kanamycin-containing medium.
quote:
How was nature structured to facilitate a complexity-increasing form of evolution driven by natural selection?
If you have a homogeneous population of replicators in an unchanging environment, there will be no selection in any particular direction. As soon as you introduce variation in the replicator population, you introduce the potential for selection. Selection is simply testing one genotype against another in a particular environment. It is a logical consequence of having different replicators in an environment.
quote:
Apart from design, sources of randomness will inform natural selection. But in that case natural selection itself becomes essentially (though perhaps not quintessentially) random.
Absolutely not. If you grow a popualtion of ampicillin-resistant and ampicillin-susceptible bacteria in ampicillin, over time you will eliminate the susceptible population. That is not random.
What CAN be random is the acquisition of resistance. For example, the classic "fluctuation test" of Luria and Delbruck demonstrated that mutations can arise spontaneously:
quote:
(taken from the Cold Spring Harbor Web Site)
In 1943 Delbruck invited Luria to do some experiments at Vanderbilt. They devised a simple experiment to test the occurrence of spontaneous acquisition of bacterial resistance to phage infection. Delbruck and Luria took samples of bacteria from small populations and from large ones. They diluted the samples and transferred them into containers of equal size, so that population numbers would be approximately equal--the only difference was the size of the parent population. They then infected the samples with phage. Normally, a certain percentage of bacteria will acquire a resistance to phage infection. Delbruck and Luria showed, however, that the variation within bacterial samples drawn from small populations was much greater than that within samples drawn from large populations. In other words, by putting the bacteria through what a population geneticist would call a bottleneck, Delbruck and Luria showed that the probability of acquiring resistance was due to an event occurring at the time of bacterial replication, not the time of separation of the populations. They had demonstrated the existence of mutation in bacteria, thus firmly establishing the phage/bacteria system as a valid one for studying the nature of the gene. This experiment, which came to be known as the "fluctuation test," was the basis of many of the experiments performed and taught by phage workers throughout the 1940s and profoundly influenced the current and subsequent generations of phage biologists.
In the above case, the acquisition of resistance is random, but the consequence of selecting for resistance is not. That, I think , is the source of Dembski's confusion.
AndyG
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 10. January 2003 19:22
The issues in this thread - the use of analogy in science (as distinguished from the pedagogical use Scott clearly intends) and displacement - have been discussed at some length several times on several boards fairly recently. Rather than repeat them here, I call readers' attention to these threads:
1. My remarks on argument from analogy here.
2. Gedanken's remarks on displacement here.
Finally, I will note that Dembski's argument in the last couple of sentences of the OP are precisely parallel to arguing that if a complex logical expression contains both a Union operator and an XOR operator, in that case the Union operator itself becomes essentially (though perhaps not quintessentially) an XOR.
RBH
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 10. January 2003 20:38
Dembski: But the deeper point is that even if on average the "more fit" survive, what determines which organisms are "more fit"?
Andy: Again, the interaction between the genotypes in a population and the environment determines what is "fit" in a particular instance. In an ampicillin-containing medium, ampicillin-resistant bacteria will be more "fit" than kanamycin-resistant bacteria.
Yes, but this is incomplete. How does the medium come to contain ampicillin?
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 10. January 2003 23:14
MG writes:
quote:
Yes, but this is incomplete. How does the medium come to contain ampicillin?
I'm sure some researcher put it there. So?
To belabor the blatantly obvious, antibiotics, and toxic compounds of all sorts, are found in nature, produced by various geological and biological processes. Surely you are not implying that we can brush off these studies of natural selection simply because the experiment, rather like most experiments, happened to involve an intelligent designer?
It might be worth noting the accidental discovery of penicillin here:
quote:
In 1928 while organizing a pile of petri dishes in the sink where he had been growing bacteria, Fleming opened each dish and examined it before dropping it into the cleaning mixture. A particular one caught his attention. Mold was growing on one of the dishes, which was supposed to happen, but he discovered that it had contaminated a staphylococcus culture and stopped the bacteria's growth. It fact, it had actually killed it. After taking a sample of the mold, he found that it was from the penicillium family. He named it penicillin and found that it was nontoxic and sufficient in treating many types of bacteria harmful to man. link
[ 10. January 2003, 23:16: Message edited by: yersinia ]
IP: Logged
|
|
Art
Member
Member # 179
|
posted 10. January 2003 23:28
Dembski: But the deeper point is that even if on average the "more fit" survive, what determines which organisms are "more fit"?
Andy: Again, the interaction between the genotypes in a population and the environment determines what is "fit" in a particular instance. In an ampicillin-containing medium, ampicillin-resistant bacteria will be more "fit" than kanamycin-resistant bacteria.
Mike: Yes, but this is incomplete. How does the medium come to contain ampicillin?
****************************************************
Antonie Van Leeuwenhoek 1999 Jan-Feb;75(1-2):81-94
Gene organization and plasticity of the beta-lactam genes in different filamentous fungi.
Gutierrez S, Fierro F, Casqueiro J, Martin JF.
Faculty of Biology, University of Leon, Spain.
The genes pcbAB, pcbC and penDE encoding enzymes that catalyze the three steps of the penicillin biosynthesis have been cloned from Penicillium chrysogenum and Aspergillus nidulans. They are located in a cluster in Penicillium chrysogenum, Penicillium notatum, Aspergillus nidulans and Penicillium nalgiovense. The three genes are clustered in chromosome I (10.4 Mb) of P. chrysogenum, in chromosome II of P. notatum (9.6 Mb) and in chromosome VI (3.0 Mb) of A. nidulans. The cluster of the penicillin biosynthetic genes is amplified in strains with high level of antibiotic production. About five to six copies of the cluster are present in the AS-P-78 strain and 11 to 14 copies in the E1 strain (an industrial isolate), whereas only one copy is present in the wild type (NRRL 1951) strain and in the low producer Wis 54-1255 strain. The amplified region in strains AS-P-78 and E1 is arranged in tandem repeats of 106.5 or 57.6-kb units, respectively. In Acremonium chrysogenum the genes involved in cephalosporin biosynthesis are separated in at least two clusters. The pcbAB and pcbC genes are linked in the so-called 'early cluster' of genes involved in the cephalosporin biosynthesis. The 'late cluster', which includes the cefEF and cefG genes, is involved in the last steps of cephalosporin biosynthesis. The 'early cluster' was located in chromosome VII (4.6 Mb) in the C10 strain and the 'late cluster' in chromosome I (2.2 Mb). Both clusters are present in a single copy in the A. chrysogenum genome, in the wild-type and in the high cephalosporin-producing C10 strains.
**********************************************
Andy's comment is more relevant to natural selection than is immediately apparent. (Until we recall the discovery of penicillin, that is.)
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 11. January 2003 09:06
The monkey typing Shakespeare experiment provides some valuable insights into the nature and effectiveness of biological design process in general and more specifically the nature and effectiveness of natural selection.
To begin, the monkey typing Shakespeare provides a reasonable estimate of the complexity or improbability of generating a Shakespearean play. Given the technique random key stroke approach the complexity is well in excess of 10^10,000. Although Shakespeare clearly used an approach other than random key strokes, the complexity or improbability, or amount of information processing required by Shakespeare to generate a play is huge. [We may not know exactly how Shakespeare wrote his play, but we do know the process involves choices or selections (select a word then given a word select spelling of the word etc. which makes it possible to show that the complexity of writing a play far exceeds the 10^10,000 level. ]
Second, we know that Shakespeare and all humans depend heavily on certain powerful front loading mechanisms to generate biological designs. A spell check is an example of a simple front loading mechanisms that would ‘select out’ any sequence of symbols which did not represent an acceptable word. We could give the monkey a series of such front loading mechanisms which would greatly reduce the time required for the monkeys to write a Shakespearean play(the time required would still by huge, but it would be shorter that the random keystroke alone method). Introducing complex front loading mechanisms might make the monkey on the typewriter closer to what actually happens in real world evolution, but such tricks would represent a major departure from current Darwinian and neo-Darwinian THEORY which does not recognize dynamic front loading mechanisms.
Third, and most telling, both Shakespeare and the monkey if given enough time, could produce complex biological designs (Shakespearean plays) without utilizing a single instance of natural selection. (not a single organism died or failed to reproduce).
We have around us abundant evidence that the production of complex biological designs involves very high levels of complexity (or very low levels of improbability). We have abundant evidence that biological systems use complex front loading processes to produce these complex designs. We have abundant evidence that these design processes and front loading processes depend almost entirely on secondary or indirect selection which is clearly a very different phenomena from whole organism ‘Natural Selection’. Despite abundant evidence to the contrary, both proponents and opponents of Darwinian theory maintain the delusion that a Natural Selection only process can or could be taken seriously as a possible explanation of evolutionary change.
William Dembski asks the question quote: "But the deeper point is that even if on average the "more fit" survive, what determines which organisms are "more fit"? How was nature structured to facilitate a complexity-increasing form of evolution driven by natural selection?". Shouldn’t we be asking instead- "Is nature structured to facilitate a complexity increasing form of evolution driven by natural selection?" Unless ‘driven by natural selection’ is interpreted to mean ‘natural selection exists and may play some minor role in some types of adaptive changes’ doesn’t the available evidence clearly show that evolutionary change is not and could not be driven by natural selection.
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 11. January 2003 12:41
I had forgotten that I was prohibited from being the first poster to respond to any given thread which is why my original posting had been deleted. The moderator was nice enough to email me the deleted text and invited me to repost it a day later. So here we go. I believe that this thread may also be relevant to another thread started by the moderator named A different sort of brainstorm where we are asked for lists of concepts that lead often to conclusions. I believe randomness as well as natural selection are two familiar terms.
I think it may be helpful to define the term random since it seems to lead to some signficant confusions.
Random mutations refer to mutations which are random with respect to the environment, that is they are not biased towards having selective advantages in any particular environment.
Is survival random? Bill argues that there are random components to survival and I surely agree but averaged over the species, is survival truely random? Is selection truely random or is it a deterministic signal with stochastic variations on top. I would argue the latter especially since random mutations alone do not seem to be as efficient as random mutations and selection in allowing an organism to adapt to its environment.
Lets look at Endler
quote:
Natural selection can be defined as a process in which:
If a population has
a. variation among individuals in some attribute or trait: Variation
b. a consistent relationship between that trait and mating ability, fertilizing ability, fecundity and or survivorship: fitness differences
c. a consistent relationship for that trait between parents and their offspring which is at least partially independent of common environmental effects: inheritance
Then:
1. the trait distribution will differ among age classes or life history stages beyond that expected from ontogeny
2. if the population is not at equilibrium then the trait distribution of all offspring in the population will be predictably different from that of all parents beyond that expected from a and c alone.
Talkorigins has a great FAQ on this topic
quote:
Replication Rules are not random in the sense that, say, Heisenberg's Principle of Uncertainty or quantum mechanics is sometimes supposed to show the fundamental randomness of reality. They are merely random with respect to natural selection. Natural selection is not random: it is the determinate result of sorting processes according to relative fitness. It is stochastic, in the sense that better engineered features can fail for reasons of probability (they may meet accidents unrelated to their fitness), but that poses no greater threat to the scientific nature of evolution than it does for, say, subatomic physics or information theory.
Another website
quote:
This criticism would be valid if evolution would depend only on random processes. But natural selection is a nonrandom process that promotes adaptation by selecting combinations that "make sense," i.e., that are useful to the organisms. The analogy of the monkeys would be more appropriate if a process existed by which, first, meaningful words would be chosen every time they appeared on the typewriter; and then we would also
have typewriters with previously selected words rather than just letters in the keys, and again there would be a process to select meaningful sentences every time they appeared in this second typewriter. If every time words such as "the," "origin," "species," and so on, appeared in the first kind of typewriter, they each became a key in the second kind of typewriter, meaningful sentences would occasionally be produced in this second
typewriter. If such sentences became incorporated into keys of a third type of typewriter, in which meaningful paragraphs were selected whenever they
appeared, it is clear that pages and even chapters "making sense" would eventually be produced.
[ 11. January 2003, 13:18: Message edited by: Frances ]
IP: Logged
|
|
Cre8ionist
Member
Member # 140
|
posted 11. January 2003 17:14
Interestingly, as noted many times before, evolutionists often slip ID into their analogies and
programs.
quote:
Eugenie's answer was that evolution isn't like randomly typing monkeys acting all alone but rather like randomly typing monkeys with someone standing at the monkey's shoulder erasing any mistakes that the monkey makes.
Now, to keep in line with Dawinian/Dawkins' orthodoxy, shouldn't this "someone" who's standing at the monkey's shoulder be blindfolded? I mean, aren't we really talking about a contrast between the seeing and the blind here?
Can the watchmaker see or not? ![[Cool]](cool.gif) .........................................Cre8
IP: Logged
|
|
Irving
Member
Member # 535
|
posted 11. January 2003 21:50
Andyg writes:
quote:
Absolutely not. If you grow a popualtion of ampicillin-resistant and ampicillin-susceptible bacteria in ampicillin, over time you will eliminate the susceptible population. That is not random.
It looks like your stacking the deck against randomness by creating a binary selection environment by reducing the selection pressures to a single environmental variable.
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 11. January 2003 23:58
Complexity grown out of random processes can be observed in myriad places in the universe. A simplistic example: rings of Saturn, in the realm of relatively simplistic physics. Even these clearly are “specified” and complex. (The specification is a circular disk field like a machine’s washer in outer appearance, but composed of many sup-particles in the form of asteroids and rocks or other material solid forms in a machine-like yet random pattern of distribution.) Rivers, underground flows mimic actions in human plumbing, clearly having possible specification. All these have, however, explanations of origin that can to varying degrees be predicted from understanding the physical relationships in nature.
What is important in all the formations I just mentioned is the “Markov chain” like sequence of events apparent in their development. One step leads to the next, even when the sequence ending in the particular formation can have a description that is highly unlikely. (Such an example is the exact characteristics of a given human, when based on a starting point of a population of humans. That a person should have a particular exact fingerprint, DNA pattern, and body image is extremely unlikely -- a point that will be important in the next part of this essay.)
Markov chain analysis of chance events based on physical constraints can be used to show that extremely unlikely configurations when taken over long sequences have very high probability relative to the just previous steps. For example body image and DNA pattern are fairly likely to develop with certain characteristics given a knowledge of the specific parents. (Children may look like parents, DNA can easily be traced to parents. Yet fingerprints are quite different. The reason is mechanisms of the first two that involve less degree of change and thus in Markov sense higher immediate probability with respect to a just prior state.)
A ready way to construct a pattern that is unlikely by 1 part in 10^150, for example, is to simply have a chain of independent probabilistic events in constructing a formation each of which is less than 0.5 probability, and in which the sequence of events is over 500 events long. (Easy to imagine a sequence of 500 parents in my example above.)
Eugenie Scott’s point may clearly be taken in terms of the Markov sense I have described here. Events are not simply unlikely or likely, rather they have those probabilities (to the extent they can be estimated) based on the point of reference being considered. The probability of a particular state can be very high when considering a Markov chain sequence, when examined from a specific near state in the chain, yet that same exact state can be extremely unlikely when considered from the viewpoint of a distant starting state in the Markov chain. And Scott’s point in the debate clearly was not to claim that the works of Shakespeare could be created “without Shakespeare or any other designing intelligence”, but rather to show the error of trying to apply a “tornado in a junkyard” model to understand such sequences of events. This is precisely my point in this essay as well.
The selection process alluded to by Scott in the form of the “erasures” from the “monkey’s shoulder” corresponds to branching patterns each of which can be followed through a Markov chain link process. So there is not only the Markov chain probability to examine, but also the process of branching of the states in the chains to consider and the effects of dealing with aggregates of populations. (Yes I realize that Dembski in some writing considers “probabilistic resources” in an attempt to reckon with this point, yet to my reading has failed to grasp the problem fully.) Erasures are simply pruning of the branching tree. There is nothing here to suggest that there are insufficient probabilistic resources to account for the successful branches in evolutionary progress.
Biologists (e.g. Dawkins) have a specific focus in which the language may not accommodate the mathematical notions that other disciplines wish to use. Specifically the notion of what is “random” and “non-random” depends greatly on the specific language of what is being discussed. For example selection by death of those organisms not well adapted clearly is a “random” event of the maladapted organisms having more random premature deaths. So a highly predictable aggregate Markov state process sequence can have at its core an aggregate of a very large number of highly random events. The biologists often are describing the aggregate behavior in models, and this can fairly be described as not “fundamentally random”, while in fact the events which compose it are indeed fundamentally random. (An excellent example is statistical thermodynamics which can easily and very accurately predict aggregate non-random behavior from fundamentally random components. Is someone going to challenge the high degree of predictability of the expansion of gas in a cylinder under specified P-V-T and energy conditions?)
The difference of perspective of statistician and biologist are very important here -- neither is wrong in their focus yet each must be understood in the descriptive context of the particular model being described. Since biologists (e.g. Dawkins) are describing the evolutionary model, that model must be understood in those terms, and any mathematics must be accommodating to what is engendered in that model. (A “tornado in a junkyard” is not an accommodating model to describe the Markov sequence behavior that such an evolutionary sequence would entail.)
Without an analysis that accommodates Markov chain like process chains, there is no basis for argument that evolutionary processes are unlikely. Where has such analysis been performed to back up such claims as Dembski has suggested in the above paragraph?
As an aside, Biologists may not be putting the evolutionary model in such Markov terminology either. But this is not done because the ability to formulate the constants would be very difficult -- rather the model is written in a much looser terminology which actually focuses on individual state transitions and short sequences of the Markov (and branching) chain. It is not a failure of biologists in evolutionary models to not construct such Markov chain models. The biologists have not made a claim of low probability in long chain sequences. Dembski appears to be doing so. Thus it is incumbent on Dembski and those who think they can develop a meaningful model to show a failure of such evolutionary descriptions to put it in those terms, and demonstrate the failure.
Clearly functionally specified structures (which can be specified after the event of their construction by evolution and thus correspond to what Dembski would call “fabrications” in claiming specified complexity) can be features of organisms in which the greater function of those structures (or improvements, possibly differential therein) imparts greater “fitness” or ability to survive in the local current environment. Who specifies which functionality imparts greater ability to survive? The answer is the local environment.
So the local environment can clearly be the source of the information. Who designs the fish eye so that it has the proper shape to properly focus, rivaling the skill of a skilled craftsman of lenses? The selection pressure of the environment of the sea, in which predators quickly finish off those fish who’s focal length is too short, too far, or distorted in one or another dimension so they aren’t as agile in locating food and danger. Paley’s designer of the fish eye’s focus is clearly “microevolution” by the processes Dawkins described -- a concept clearly demonstrated and accepted even by ID promoters an creationists alike.
It is not clear which aspect is being questioned (randomness of the “generate” process of mutation and combinational effects, or “test” process of natural selection). But the “generate and test” procedure is clearly one that has been demonstrated to exhibit aspects of intelligent behavior in trial after trial of artificial intelligence research. Generate and test is a common “Gestalt” in AI search methodology, and one of the latest instantiations is the “Genetic Algorithm” which mimics evolutionary processes to perform intelligent behavior. So if there is not intelligence in the random “generate” portion, and the “test” in a GA simulation of actual nature is based on accurate representation of conditions of nature, where is the “intelligence” coming from -- the answer is clearly the fundamental characteristics of nature itself and its relationships.
Dembski mentions the question of infinite regress, and I think this is important point. So (as I have pointed out elsewhere) if we regress the question to the source of the structure of nature itself, indeed we have an interesting philosophical problem. (Besides the links to our previous statements provided by RBH above, this may be of interest.) I don’t think that science can answer that. And if the wondrous structure of nature around us (and awe thereof) leads to suggestion of intelligent design of the universe and its properties -- I certainly have no argument. Is there an argument here with the question of the accuracy of the theories of evolution?
By the way, I realize that the example in my first paragraph of the rings of Saturn would not be considered “complex” according to the very technical definition used by Dembski -- precisely because in the Markov chain sequence of events we can (most likely) create a model in which we understand the local constructive behavior. (Now physicists may not have done so yet -- and thus the theory of the structure of the rings of Saturn may be “incomplete”, as much of it remains unexplained -- a common claim by ID promoters as a criticism of evolutionary theory as well. Does this mean that the rings of Saturn are indeed “complex” and “specified”, because the theory is not complete?) So now use the very technical definition of complexity when viewing the evolutionary scenario. With the model suggested by Scott, life is simply not “complex” according to Dembski’s very technical definition, taken in light of Markov chain descriptions or modifications to his ideas to accurately describe the requirements I have suggested here. If you don’t like that definition, I am sorry, but science is a discipline in which one defines terms and uses them consistently. If the word is suggestive of something that misleads the reader, perhaps the word should be changed -- should we in this case change the word “complexity” to another term in Dembski analysis? I just didn’t want to give someone an easy out to dispute my understanding of ID concepts just from my first paragraph.
With the complexity in Paley’s design of the fish eye (referring to the focusing aspect only in this instance) being explained by environmental pressure in “microevolutionary” processes, we now must consider the longer chains of just the same Markov sequences. The important notion of a Markov chain is that there is no limit to the number of states that can be concatenated in sequence. Likewise, where is the limit on microevolution? The long Markov chain is explained by understanding the short chain’s aggregate behavior. Macroevolution is explained by understanding microevolution.
[ 12. January 2003, 02:44: Message edited by: gedanken ]
IP: Logged
|
|
Cre8ionist
Member
Member # 140
|
posted 12. January 2003 10:41
Hi gedanken,
quote:
So if there is not intelligence in the random “generate” portion, and
the “test” in a GA simulation of actual nature is based on accurate representation of
conditions of nature, where is the “intelligence” coming from -- the answer is clearly the
fundamental characteristics of nature itself and its relationships.
I'm not quite sure if this was tongue-n-cheek or not. Do you then claim that Mother Nature has an IQ as well? What might it register?
Does this mean that your answer to Dembski's question would be that it is Mother Nature who's whispering in natural selection's ear?
And if so, how big can her vocabulary be?
To return to the typing analogy for a moment:
When the typist gets to "METH" (of METHINKS IT IS LIKE A WEASEL) , isn't Mother Nature silent, never able to utter the final form of the phrase? For "METHI" would qualify as an unselectable step. And even if you expanded your vocabulary to include more steps, eventually you still might reach an unusable string. After all, there are more harmful mutations than beneficial. So what would she whisper in such a case?
This is why Behe's irreducible complexity is such a goldmine for research.
Biological machines which have potentially unselectable steps in their origin paths can invalidate nD. In other words, Mother Nature's vocabulary may not be large enough to do the job of building a flagellum etc.............................Cre8
[ 12. January 2003, 10:54: Message edited by: Cre8ionist ]
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 12. January 2003 11:27
Thanks for the interesting questions, Cre8ionist. I was completely serious, but I am not sure if you were.
In anthropomorphizing “mother nature” you set up a scenario in which an agent can be considered detached from the system of nature itself. That of course was not my intention. (But the problem of infinite regress of course does not go away, but that has always been a philosophical problem of ultimate creation as opposed to proximate cause.)
No separate agent is “whispering” to the mechanism of natural selection. The fish that can’t see straight doesn’t get caught by a third party, it gets caught by the predator that can see.
Now how do we rank the degree of “intelligence” when we don’t even have a common language for discussion of creative capacity of nature and of human beings? By one standard we note that humans have hardly even been able to create life at all (what was it, a polio virus or something?) So using that as a standard then nature’s “IQ” would be incredibly high. And of course according to most theories of evolution our intelligence develops from the characteristics of nature, and thus nature must have vastly superior creative abilities to our own, since the whole is clearly greater than the sum of its parts and humans are clearly part of nature. (Or do you choose to separate human beings as not part of nature?)
Dawkins “Methinks…” example was of course supposed to be an example to demonstrate a refined and abstracted principle, and not to be a valid example of the larger picture of evolution.
Now as to the “vocabulary of nature”: We must separate the issue of what are reachable states in genetic structure by evolutionary pathways from what are possible states that can exist in nature of such genetic structure. To (for the moment) give some latitude to the claims of “irreducible complexity”, we would ask separate questions as to whether a flagellum can exist in nature, from the question of whether the mechanisms of evolutionary change could reach the genetic state of the flagellum from an ancestor that didn’t exhibit those descriptive properties. If it can exist in nature, then clearly it is in the vocabulary of nature (as opposed to the question of whether it is among the “written sentences” of evolution).
The answer is clearly that bacteria which exist in nature are seen to regularly reproduce, bacteria that have the flagellum structure. So nature clearly produces the flagellum, you can put out food and starter bacterial and wind up with a very large number of flagella created in nature. So the flagellum is clearly part of the “vocabulary” of nature.
Now the issue of “irreducability” of the bacterial flagellum and its possible significance to evolutionary theory is another discussion. I think there are other threads that discuss that, and I will not continue here. See Ken Miller’s article on the flagellum. Perhaps someone can name threads devoted to that subject.
I almost forgot, the question of more failures than successes in mutations. One does not need a large fraction of mutations to be successful for those few that are to predominate. See Talk Origins discussion of this subject. Once again not a subject I suggest for this thread, as I am sure the topic has been discussed at length in other threads.
[ 12. January 2003, 12:04: Message edited by: gedanken ]
IP: Logged
|
|
|
Printer-friendly view of this topic