Is “Directed Evolution” Darwinian? [with addendum]

_{William Dembski

April 11, 2007

Darwinism, Evolution, Intelligent Design}

Share: Facebook; Twitter; LinkedIn; Flipboard; Print; Email

I posted a reference the other day to a peer-reviewed paper by two Finnish ID-supporters that I claimed supported ID. The paper highlighted that evolutionary methods work to the degree that they are directed. As is typical with our detractors, whenever a pro-ID paper by pro-ID scientists comes out in a peer-reviewed biology journal, they try their best to show that it doesn’t actually support ID. An example is the following post at PT by Steve Reuland:

pandasthumb.org…the_proid_paper

In reading Reuland’s critique, try to keep track of “rational design,” “directed evolution,” and “Darwinian methods.” Reuland conflates the last two. In so doing, Reuland completely misses the boat. So let me spell it out: DIRECTED EVOLUTION IS NON-DARWINIAN. DARWINIAN EVOLUTION IS NON-DIRECTED. I’ve been saying this now for close to a decade (see ch. 4 of my book No Free Lunch). Just because the word “evolution” is used doesn’t mean that homage is being paid to Darwin. “Directed evolution” properly falls under ID.

[Steve Reuland, commenting at the Panda’s Thumb on this post, claims that I’ve misrepresented him and the paper. If he but were to read the paper closely, he would find that it distinguishes between Darwinian evolution as an “inspiration” to directed search and “Darwinian blind search” as inherently limited. Darwinian evolution, which is blind, is the inspiration for evolutionary computing, which employs well-crafted fitness landscapes to achieve ends and therefore is not blind — and therefore is properly a branch of ID and non-Darwinian. Yes, we’re playing a turf war here. But it will not do to have Darwin discard teleology and then to claim teleological processes as Darwinian. This is an abuse of language. Leisola and Turunen skirted the edge yet nuanced their views adequately; but Reuland is guilty of it.]

Comments

Someone mentioned in a earlier comment that whereas WEASEL totally misses the point of RM+NS (I quite agree, you've all done quite enough to demolish it), a program that took an already meaningful phrase and had it mutate and select for new, different but meaningful phrases could be interesting. Such a program require not only that mutated letters would produce not only real words, but words that go together syntactically and semantically. And without any specific end target. Has anyone ever worked on a program that does this? It would be very interesting.motthew_{April 12, 2007
April
04
Apr
12
12
2007
09:36 AM
9
09
36
AM
PDT}

I think this all serves to highlight something I've long considered: Agent causation behind the universe or broad natural forces is difficult to prove, due to the scope of those forces and the problem of discerning intention. On the other hand, random/purposeless activity is just as hard (or perhaps, even harder - we know for certain that agent causation exists on some levels) to prove, largely for the same reason: If you're not certain what would have been intended to begin with, how can you assert an action is unintentional? It all comes down to philosophy in the end. But here's the problem from this ID-sympathetic person's perspective: 1) Both of these views clearly fall more into the realm of philosophy. Darwinists insist one view is scientific and essentially fact, and the other is not. 2) A person can "do good science" while holding either one of these views, because understanding mechanism can be done apart from discerning intention and ultimate causation. Darwinists insist that even entertaining the thought of intelligent order will result in bad or even no science. I think it's clear that Darwinists are wrong on both 1 and 2, and the debate has never really been about 'protecting science', but protecting a worldview that requires an exclusive claim to science for morale reasons.nullasalus_{April 12, 2007
April
04
Apr
12
12
2007
09:34 AM
9
09
34
AM
PDT}

Atom, Regarding comment #46: The distinction your making between natural and artificial selection is interesting, but lacking. Let me make sure I understand you correctly: You're saying that in the artificial scenario, selection is a separate and earlier action to differential reproduction. Whereas, in the natural scenario, selection and differential reproduction are one and the same. This idea makes sense if we're only considering the production of a single generation. But consider it as an iterative process. Once the offspring are born, a subset of them will die before they reproduce themselves. Thus, nature prevents those with "undesirable" traits from reproducing, just as a human prevents dogs with undesirable traits from reproducing in a breeding program. But that doesn't make artificial and natural selection the same thing. The force of artificial selection is much stronger: a person takes one male and one female that are most likely to give the desired result. Natural selection allows all the population to freely mix, severely diluting the most desirable traits. At the very least, this means that natural selection is a very slow and inefficient process, and possibly so inefficient that is not able to do much beyond certain limits. How to discover those limits, well.. we've been debating that for almost two hundred years.motthew_{April 12, 2007
April
04
Apr
12
12
2007
09:28 AM
9
09
28
AM
PDT}

Atom, If you want to argue that in artificial selection the use of the word selection is more appropriate that is fine. But in natural selection, conditions in the environment act to affect the passing on of certain alleles so Darwin used the term selection to contrast this process with the obvious one everyone understood which was artificial selection. If you do not think that Darwin's use of the term "selection" is appropriate then it is your prerogative to do so. Darwin was rhetorcist supreme and his choice of words had their desired effect. The process happens and you can call it what you wish but everyone knows the term, natural selection. Since the process is real and obvious, your objection is coming off more as a quibble than one of substance. I personally don't care what the name is and if it is an inappropriate use of the term "selection." What ever you want to call it, the process has trivial effects and that is what should be hammered. Quibbling with the actual meaning of the term "selection" itself based on its dictionary denotation distracts from the main issue and I am not sure what it gets anyone.jerry_{April 12, 2007
April
04
Apr
12
12
2007
09:21 AM
9
09
21
AM
PDT}

"But to be fair, we must remember that differential reproduction depends on the Ã¢â‚¬Å“interactionÃ¢â‚¬Â between the organism and the environment, so the environment has some role in the equation, although passive."
I agree. It does play a role which is passive; it is the backdrop. The "interaction" is what takes place as the replicators out-replicate each other, the environmental constriants providing the backdrop. The end result is the same as the process: differential reproduction. Therefore, I still hold that NS is not causal.Atom_{April 12, 2007
April
04
Apr
12
12
2007
09:21 AM
9
09
21
AM
PDT}

Atom: "It is backwards: differential reproduction is the Ã¢â‚¬Å“causeÃ¢â‚¬Â of what we label selection, which is just describing the end result of the process, which is differential reproduction. Again, differential reproduction is the Ã¢â‚¬Å“causeÃ¢â‚¬Â of differential reproduction." I think you are correct in you statements. That's what NS is. But to be fair, we must remember that differential reproduction depends on the "interaction" between the organism and the environment, so the environment has some role in the equation, although passive.gpuccio_{April 12, 2007
April
04
Apr
12
12
2007
09:13 AM
9
09
13
AM
PDT}

jerry, your post is a very clear and balanced evaluation of the problem. I have just one comment: we have two kinds of alleged examples of NS, and they are very different in their meaning. In the first kind, we know the specific mutation we are talking about, like in the much discussed scenarios of sickle cell disease, antibiotic resistance, OP resistance in mosquitos. All the cases in this group seem to share three constant characteristics: a) The mutation is a single, simple mutation, ususally a single bp variation. We are therefore in the legitimate field os spontaneous random mutations, occurring with reasonable frequency. b) The mutation involves, invariably, some loss of information in the protein which has mutated, more or less serious. c) The mutation may involve some indirect advantage in specific environmental situations, ususally related to the modification-damage of the interested protein (never to the appearance of a new function), and so it can be "selected" in that context with the mechanisms you described: in the end, the prevalence of the mutated allele may be higher than expected. In the second kind of examples that you cite, instead, (finches, moth, bears, etc.) we know nothing (at least, to my understanding) of the molecular basis of the supposed mutation. That's a very important difference, in my opinion, because we can have no idea of what is really happening, or of the mechanisms involved. All of these examples seem to reflect some simple, positive adaptation to the environment, and so they could be considered a little bit "higher" than the examples in the first group, which don't even deserve the name of "adaptations". But my problem is: if we don't know the molecular mechanism, how can we be sure that it is RM + NS? As we are no more in the naif times of Darwin, I think that we can define something a "mutation" only if we are aware of the specific underlying genomic variation: otherwise, how can we know that it is not a true "adaptation" (in a Lamarckian sense), and not a darwinian phenomenon? We have seen a great revival of Lamarckian concepts in a new form in the last few years, including all we know (and especially all we still don't know) about epigenetic inheritance. So, to sum up, you are perfectly right: the few well known examples of RM + NS are at best responsible of trivial variation inside the species. But still, there is a difference between those examples in which the molecular basis is known, and our inferences are more detailed and justified (and yet, even in these cases, not all is necessarily as it is supposed to be, as we have discussed elsewhere), and those other examples where everything is highly hypothetical (finches, etc), and whose fortune in the darwinian community seems to be more a sentimental and cultural question (which I can understand: depriving a darwinist of his finches would certainly be pure cruelty!) than a scientific one. And, finally, a note about genetic drift: why is everybody in the darwinian community so excited about that? It seems obvious that it is a purely random phenomenon. Either it happens or not, either it is relevant or not, it is totally irrelevant to the problem of information buildins: at best, it is a random variation of random variation, adding nothing to the power of RM. In other words, we know existing information can vary in a certain number of random ways: be it single nucleotide substitution, single nucleotide deletion or addition, microdeletion, macrodeletion, chromosomal alterations, gene duplication, genetic drift or whatever else, who cares? What is the difference?gpuccio_{April 12, 2007
April
04
Apr
12
12
2007
09:08 AM
9
09
08
AM
PDT}

For the sake of everyone (including myself) let's take a step back and again contrast Artificial and Natural Selection. In Artificial Selection you have a group of organisms and an agent. Before any reproduction occurs, the agent will select which organisms will be allowed to reproduce. (Like a dog breeder does.) This is a definite action. Thus, the selection takes place prior to any reproduction occuring. After the selection takes place, the selected organisms are allowed to reproduce, which is the differential reproduction. It is clear that the artificial selection is an actual cause that exists prior to differential reproduction occuring and causes the differential reproduction. Now let us replace the agent with nature. Take a group of organisms that have not reproduced yet. Nature has a set of constraints that exist there as a back-drop. Nature has done no selection yet, since no reproduction has yet occurred. Now the organisms are allowed to reproduce, some producing more than others. This is the differential reproduction (or differential allele passing), which is described as the result of some organisms reproducing more than others. Those organisms that reproduced more are the ones that are "selected." When and how did the "selection" take place in this second scenario? The act of reproducing more IS the act of selecting. Contrast this to the first scenario, where the act of selection is separate from and temporally prior to the act of differential reproduction. Organisms are "selected" as they out-reproduce their neighbors; in other words, the organisms "select" themselves. But if this act of differential reproduction is where the selection occurs, how can selection be said to be the cause of differential reproduction? It is backwards: differential reproduction is the "cause" of what we label selection, which is just describing the end result of the process, which is differential reproduction. Again, differential reproduction is the "cause" of differential reproduction.Atom_{April 12, 2007
April
04
Apr
12
12
2007
08:43 AM
8
08
43
AM
PDT}

Hey Jerry, Thank you for your time as well and the discussion. Now, I know that differential reproduction takes place; I never denied this. I also know that those organisms that have traits which allow them to reproduce more (differential reproduction) will be the ones that produce more (tautology, since we identify and assess the fitness of those traits by actual reproductive success), and therefore over time there will be more of those types of organisms around (another obvious statement, since the survivors survive). You wrote:
Natural selection is a process that produces differential offspring of certain alleles.
Are you saying that Natural Selection "causes" differential allele frequencies? If so, how is that any different than the process itself, which is just the description of organisms mutating and trying out their different configurations agianst a fitness backdrop of constraints by out-reproducing each other? NS is or at very least includes differential reproduction (differential allele passing) as a prerequisite for the process to occur. How can you then say this differential reproduction is an "effect" of NS?Atom_{April 12, 2007
April
04
Apr
12
12
2007
08:24 AM
8
08
24
AM
PDT}

Atom, Natural selection is a process that produces differential offspring of certain alleles. Why? Sometimes it is not obvious but sometimes it is. When an allele or combination of them make the organism more likely to survive to reproduce then the term survival of the fittest applies. Genetic drift is the major cause for allele frequency changes in the population so this is not natural selection. But there are cases where some allele combinations lead to differential reproduction because it leads to success for the organism. It is not circular because it tries to identify the relevancy of an allele or combination of them for survival and reproduction variations. Whether this happens much is another story. We have had these discussions here recently, namely, the sickle cell debate and it seems there are not too many obvious documented cases. Darwin's finches and the peppered moth (if it happened) are trivial cases of NS which probably did not involve any new allele creation. Polar bears and grizzly bears can mate but white fur is probably a better survival allele on the polar ices and a brown color better for forrest areas. Again a simple example. Given that NS can happen, the real story is not that it can happen but that it is really a trivial process and yet the Darwinists must have it as their magic process. It is the non random part of their theory but as I said before it is "ho-hum." It is embarrassing that they embrace it so fervently.jerry_{April 12, 2007
April
04
Apr
12
12
2007
06:41 AM
6
06
41
AM
PDT}

Matthew Tan, Why don't you read what Dembski actually says. He has several things on his other site. http://www.designinference.com/ Read Dembski's two very content full essays he wrote for the Dover Trial. Nearly everything written about ID and Dembski, Behe etc is false or distortions of their positions. Read what Dembski and Behe say and then what Darwinists write about them. I have never seen an honest Darwinist. According to a Darwinist if you believe in God and believe he had something to do with the world, you are a creationist.jerry_{April 12, 2007
April
04
Apr
12
12
2007
06:25 AM
6
06
25
AM
PDT}

Did Dr. William Dembski say this? This is from a Amazon.com's book reviewer: http://www.amazon.com/gp/product/0393050904/103-9385963-4546267?ie=UTF8&tag=origiresou-20&linkCode=xm2&camp=1789&creativeASIN=0393050904 "The creationist community, on the other hand, is both narrow and narrow-minded. Many of its most prominent advocates, for example, including William Dembski, have publicly proclaimed that they would remain committed to creationism regardless of what the evidence showed."MatthewTan_{April 12, 2007
April
04
Apr
12
12
2007
01:16 AM
1
01
16
AM
PDT}

DaveScot said:
I usually just grab the system timer count in microseconds and use one or more of least significant digits as my random number.
I imagine this would work just fine, as long you weren't grabbing successive numbers in a loop, where I would worry that the timing interval between iterations might stabilize. I haven't tested it in this fashion, I've usually just relied on system provided random number functions. (For integer random numbers, call FRand() and multiply by the desired random number sample: FRand() * 100 for a 1 to 100 or 0 to 99 result)Apollos_{April 12, 2007
April
04
Apr
12
12
2007
12:30 AM
12
12
30
AM
PDT}

Of course agency can affect natural processes so here we have a force outside of normal natural processes that can affect outcomes. If we assume the agency has free will, that is it is not determined, then we can say we have two forces not the three we usually postulate. Namely, law and agency. But if there is no free will, then there may be only law.
Thanks for this, Jerry. It seems reasonable to me that free will is indeed a factor, after all I'm the one writing this now, not purposeless forces. I guess my initial thought wasn't too far off the mark if I'm right about free will: randomness may be an artifact of perceptual limitations, an "optical illusion" of sorts. Predictability may not always be within our ability to reason, but it may not ever be unreasonable.Apollos_{April 11, 2007
April
04
Apr
11
11
2007
11:12 PM
11
11
12
PM
PDT}

As for fitness landscapes, I understand the concept. They are contraints that replicators must fulfill to reproduce more or less offspring, depending on how well they fulfill these constraints. (Assuming no "noise" from chance events.) But notice, it is the organism itself which either exploits these constraints or does not; nature does nothing but is merely the back-drop for these exploits. And of course, by exploiting the constraints, I mean out-reproducing the other replicators. Yet again we return full circle to differential reproduction...Atom_{April 11, 2007
April
04
Apr
11
11
2007
10:58 PM
10
10
58
PM
PDT}

gpuccio, pleased to meet you as well. I agree, NS does describe "something", there is something going on. And I also agree it is not a "force" as some overly enthusiastic supporters claim. great_ape, thank you for taking your time to dissect my ideas. I appreciate your courteous interaction on the topic. I hope not to bore you or waste your time.
Natural selection includes the entire *process* that yields differential reproduction, not just differential reproduction itself. The prerequisites for the algorithm of natural selection are usually something like as follows: a) excess offspring b)the presence of heritable variation c) differential fitness / reproductive success based on that heritable variation.
Your answer is telling. You said that one prerequisite for Natural Selection was "c) differential fitness / reproductive success based on that heritable variation" also known as differential reproduction. (How else can "fitness" be measured? An organism is only as "fit" as the extra fraction of offspring it produces.) But you also said that the end result is differential reproduction: "...the entire *process* that yields differential reproduction" So, if I understand you and Darwinists correctly: Natural Selection is a process that requires differential reproduction as a prerequisite to yield differential reproduction. You see my problem with that? To me it is like saying that the loss of hearing causes deafness. You conflate cause with effect, and show that Natural Selection has no causal power. Maybe you can be clearer, what does Natural Selection cause?Atom_{April 11, 2007
April
04
Apr
11
11
2007
10:47 PM
10
10
47
PM
PDT}

Dave, thanks for the response.
No one knows if there is such a thing as truly random. The answer to that would spell the difference between a deterministic and non-deterministic universe.
This cuts right to the heart of my inquiry. I'll check out the Scripps research link as well.Apollos_{April 11, 2007
April
04
Apr
11
11
2007
10:46 PM
10
10
46
PM
PDT}

Apollos, The concept is indeed of importance and fits this topic as much as a discussion of natural selection does. In another recent post, gpuccio, discusses how what may appear random is the result of a multitude of forces. Since we are unable to understand how they interact the distribution of results are impossible to predict and are best described by a probability distribution. But each result is determined and could be predicted if we only understood all the forces involved. Now I do not understand quantum mechanics but it is based on random processes and for most of his life or even till he died, Einstein really believed that they were not random but determined and that we just did not understand what caused the randomness. Some have said that they really are random and if you could rewind the clock you would get a different answer which would not be true under a determined universe. So if there is really no randomness but just a lack of knowledge on how an individual result came about, then the entire natural world is determined. Under this scenario, life if it came about by natural causes would be the result of law and not chance. So if someone set up the initial conditions and life appeared it was a result of the initial conditions and the laws of the universe. Of course agency can affect natural processes so here we have a force outside of normal natural processes that can affect outcomes. If we assume the agency has free will, that is it is not determined, then we can say we have two forces not the three we usually postulate. Namely, law and agency. But if there is no free will, then there may be only law. I am rambling on but I think this is an essential question that we throw around without really thinking about it very much. Others may want to comment on the idea that randomness is really not random.jerry_{April 11, 2007
April
04
Apr
11
11
2007
10:26 PM
10
10
26
PM
PDT}

Appolos Random loosely means unpredictable. To be more precise truly random means there is not and never will be a means of prediction. No one knows if there is such a thing as truly random. The answer to that would spell the difference between a deterministic and non-deterministic universe. Pseudo-random is often used in computer science to describe algorithmic methods of producing random numbers where predictability exists only if you know or can deduce the algorithm. Evolution pundits tend to abuse the term by implying that random mutation is truly random (with respect to whether or not any given mutation is advantageous) - a totally blind process. This is not proven but rather taken as a matter of faith in The Church of Saint Charles Darwin. It may very well not be blind at all and lots of recent research indicates it isn't. The case we usually note here is the Scripps research which found that bacteria control the rate of mutation and step on the gas pedal when there are toxins in the environment. Here's an analogous situation. An accomplished dart player doesn't always hit the bullseye but few people would characterize his attempts as random or blind. More and more what was thought to be blind chance is not at all blind but merely imperfect as the dart player is imperfect but not blind. The bacterial mechanism reported by Scripps is not blind. It knows there's a target and when there's a reason to hit it. It just doesn't know exactly where the bullseye is located. Designers call what the bacteria is doing the "shotgun" method. A shotgun blast contains many projectiles that cover a wide area compared to a rifle shot. It's less efficient but your aim doesn't have to be as good. By the way, in generating random numbers in personal computer applications where reproducibility isnt' a requirement and consecutive samples are separated by many milliseconds or more I usually just grab the system timer count in microseconds and use one or more of least significant digits as my random number. There are a lot of asynchronous events of varying duration happening in a typical PC so the precise number of elapsed microseconds from one sample to the next is sufficiently unpredictable. If I need a lot of samples quickly I'll usually grab the timer count and use that as the seed for an algorithmic pseudo-random number generator. That's more or less de rigueur in game programming. For higher quality random numbers, again where reproducibility isn't a requirement, a convenient source is a sound card. Crank up the amplification on an open MIC input and use one or more of the least significant digits from consecutive samples of the A/D convertor. In that situation you're reading semiconductor noise in the amplifier. DaveScot_{April 11, 2007
April
04
Apr
11
11
2007
10:21 PM
10
10
21
PM
PDT}

Dana, Atom: perhaps I would not describe myself as "a socialist of sorts", but I am certainly not right-wing. And I definitely am an ID advocate. So, good to meet ya both.gpuccio_{April 11, 2007
April
04
Apr
11
11
2007
09:04 PM
9
09
04
PM
PDT}

great_ape: if your discussion is strictly about the "causal" word, I think we can find a compromise specifying better the terminology. I think what Atom means (and in a sense it is a good point) is that it is wrong to conceive NS as an actin (and theresore "causal") force. On that I agree. NS is not a force, in any sense that a force is defined in natural sciences. But you are right that, is something called NS happens, it must have a cause. Assuming for an instant that NS happens according to darwinian paradigm, then the "cause" of what Atom correctly calls "the result" can be best described as an interaction between the reproductive ability of the organism (depending on the total biological information it owns) and the resources available in the environment (and obviously the obstacles and limitations inherent in the same environment). So, the environment (call it nature, landscape, or any other name) is only part of the equation. The other part is the active biological information in the organism. And there is no force in action, only a process which can, but not easily or simply, be described as the result of complex necessity laws shaping an interaction. But the real proble, beyond the "causality" problem, is that of information. Excuse me if I am repetitive, but that's the real question. You say: "By drawing a very straightforward analogy, humans play the same role in artificial selection that the environment plays in natural selection." Well, let's analyze better the differences. It is not only a problem of different agents (landscape against intelligent agent). We can view it as a problem of information. By definition, the "landscape" has no specific information neither about the organism's already existing information nor about the new information necessary to have a "new" organism (in other words, the necessary variation to express a new function. The only role of "landscape" is a blind one: allowing or denying resources of which it knows nothing, creating or subtracting obstacles of which it knows nothing. In that sense, the landscape is a totally random factor. It certainly obeys the laws of physics and necessity, but there is no explicit link between those laws and the organism. The landscape is, after all, a "blind watchmaker", even for Dawkins. For me, it is blind, period. The real "teleological" element, the only non random element in the NS equation, is not in the environment, but in the organism: the organism is certainly not random, and the information it already owns is its non random heritage. In the case of intelligent causality, instead, the scenario is completely different. The material cause of the variation can or not be the same (random variation, see the weasel example and my previous post about antibody maturation). But intelligent agents can use, and indeed use, other "causes" of variation: a programmer does not write his code through random variation of the letters and selection of the right ones (although he could). A programmer usually directly "types" the correct letters. And even if the intelligent agent decides to use random variation as the means of modification, he will always have to exercise intelligent causation, through any available method, to select the result. Because, you see, that's the only way that he can impart the new information he already knows to the existing pattern. Tha fact is, darwinian model cannot do that. In darwinian model you have three components: RM, landscape, and existin organism. Of these, two are completely random, in respect to the "new" complex, functional information which should emerge. They are RM and landscape. Please note that landscape is certainly random in respect to the new information because it has no knowledge of what that new information must be, even if it can be a limiting or enhancing factor in the interaction with the "new" organism which will have the "new" information" (with its inherent function). Let's pretend, to be clear, that the new information is an oredring algorythm. Let's pretend that the landscape can, in an interaction with an organism, positively enhance the new ordering function. Still, the landscape is blind, so it has no idea of how an ordering algorythm should be written to get that function. In other words, it has no information about the information which has to be created, even if it can have a role in the interaction once the new information is created. So, we have random mutation, random environment. But yes, we have a non random factor: it is the information alredy existing organism. But, that's exactly the information we want to "increse". In our example. the existing organism does not include on oredering algorythm. That's exactly what it should acquire, to become a new organism. And, making exception for small variations, usually not functional, or only indirectly functional, like those examples we have often discussed, and in absence of any functional coninuity of ordering algorythms (and, believe me, of any more than basic function in the world), there is only one way the old organism can "learn" the new algorythm, be it an ordering algorythm or a new enzyme, or anything else: it has to come from outside. The point is, it is not necessary that an intelligent agent be the direct cause of that information imparting. He can also be the indirect cause, in the sense that the information can be written somnewhere (always by an intelligent agent) and, at some point in time, transferred to the organism. But the point is, there is no way in the world of having CSI unless an intelligent agent has created it. So, you see, we can call a selection "intelligent selection" if the information (in the sense of CSI, that is prespecified, or compressible, or functional information, that is a rare, very rare island in the ocean of possibilities) is already known to the system and can therefore be "cloned" on a new "hardware". That's the only true "selection" I know. The agent, or the information he has already created, "selects" something which is already known, already available. So, in that context, the only real cause of the new information is the existing information, although many intermediate causes of different nature (random, not random) can be used. We are, perfectly, in the "weasel" paradigm (thank you, Dawkins). We are in the paradigm of ID.gpuccio_{April 11, 2007
April
04
Apr
11
11
2007
09:02 PM
9
09
02
PM
PDT}

Would anyone care to comment on the concept of randomness? It comes up often (over 20 uses in this thread), and I sometimes get tripped up by whether the concept is concretely real, or a useful perceptual limitation. For instance, if I take a digital stopwatch that displays thousandths of a second -- each time I push "stop" I'll get an apparently random number at the third decimal place. This might prove useful, but isn't truly random, as the reading is determined by exactly when I press the stop button. Using random numbers in a computer program is as easy as asking for one; but building a random number generator (a Pseudo Random Number Generator) requires pre-established lists of "random" numbers, such as from decimal places in Pi. These might be indexed based on the sum of the last three decimal places from a high-resolution timer (similar to the stopwatch method above). It is illusory randomness, and can have the benefit of being predictably repeatable. Testing "background" noise is another method, but randomness would seem to be illusory there as well. Although not necessarily predictable in any meaningful sense, urban noise for instance, would be determined by cars, voices, fans, doors opening and closing, airplanes, etc, each the product of some sort of deliberate action. A coin flip is apparently random, but if it lands in the hand at least, is determined by the force applied to spin it, the distance it travels, etc. I guess I'm wondering about the philosophy of the nature of randomness, whether it objectively exists. The concept is useful in a variety of applications, but is there room to interpret it as a convenience of perceptual shortcomings? If this is too off-topic, please disregard.Apollos_{April 11, 2007
April
04
Apr
11
11
2007
05:52 PM
5
05
52
PM
PDT}

I've never seen the "weasel" program run, but I think I understand what it attempts to show. If I'm correct, it demonstrates exactly the opposite of what it claims, making it a perfect example of framing and no more valuable to knowledge of the subject than the video demonstrating evolution (and may instead subtract from the sum total of human knowledge, ironically paralleling that RM does little more than damage existing information). My understanding of the "weasel" program: A string of apparently random characters magically transforms into a recognizable, ordered statement. I'll use my own text for the example. A string: "uwocaeqnrbw pkmvxz" ...steadily becomes... "intelligent design" ...apparently by random selection. The fact is, that the target pattern must already exist in the code of the program (or some other data source) exactly as it needs to appear, letter for letter. A pseudo code example might look like this: templateStr[] = "Intelligent Design"; randomStr[] = "uwocaeqnrbw pkmvxz"; foreach( character chr ) in randomStr[] { while( randomStr[chr] <> templateStr[chr] ) { randomStr[chr] = GenerateRandChar(a...z); display( randomStr[] ); } } For each index in the random string, a random character is generated until it matches the corresponding character in the template string. Each time a random character is generated, the random string is displayed, completing the illusion of random activity "generating" specified information. It is misdirection, prevarication. The program takes a deliberate path around the truth. Where it could simply { display( templateStr[] ); }, it instead jumps through hoops to "simulate" a blind search for a meaningful pattern. Perhaps there is more going on in the weasel program, but it most likely amounts to the same. The target pattern must be a template, and every successive "generation" would have to compare itself against that template. If a program could change a meaningfu statement one letter at a time -- with each generation conveying specific meaning -- and transform it into a new statement with different but equally specific meaning, I wuld be more impressed, but that would not skirt the design issue.Apollos_{April 11, 2007
April
04
Apr
11
11
2007
04:30 PM
4
04
30
PM
PDT}

"Artificial selection is teleological and intelligent directed. So they are very different." --jerry I'm not claiming they are the same; I agree that the presence/absence of intelligence is a crucial distinction. However, here I am simply claiming that the two manners of selection are not different in the sense that one is causal an the other is not causal. Atom, I think the weak point of your argument is illuminated by the quote from Darwin you have provided. To put it in your terminology, nature/environment (including other organisms) are acting as "force A". There is a reason that Darwin describes it in the fashion that he does; there is a sort of agency at work in his opinion; it's just not an intelligent one. Natural selection includes the entire *process* that yields differential reproduction, not just differential reproduction itself. The prerequisites for the algorithm of natural selection are usually something like as follows: a) excess offspring b)the presence of heritable variation c) differential fitness / reproductive success based on that heritable variation. Included in the notion of "differential fitness" and, similarly, "differential reproduction" is the concept of a fitness landscape. It's a sort of equation, with various parameters, that determines how successful a given individual will be based on its genotype/phenotype. My main point here is that this fitness equation is *determined* by the environment of the organism in very much the same way as a human, during artificial selection, determines the "fitness" of individual traits by deciding which animals to breed. It's strange that you're willing to extend "artificial selection" beyond the confines of the results to see the "force" at work, but you are unwilling to consider natural selection beyond the "result" of differential reproduction. Now, I agree with Jerry, that there is a large difference between intelligence being involved and nonintelligent processes, but I don't think this has anything to do with "causality" as such. Perhaps you are instead getting at the fact that one scenario has "intelligent agency" and the other does not? I think the fact that you naturally envision nature as an agent of action is why you're unwilling to assign nature/environment the very same role.great_ape_{April 11, 2007
April
04
Apr
11
11
2007
04:07 PM
4
04
07
PM
PDT}

When dealing with the writings of such complex people, it is often times difficult to sum up what the heck they are actually talking about. The good thing about Tom Wright is that he tends to be a straight shooter.DanaMcgee_{April 11, 2007
April
04
Apr
11
11
2007
03:59 PM
3
03
59
PM
PDT}

Hey Atom, good to meet ya. Yeah it's not as big a contradiction as it first seems. I know Bishop Tom Wright (Anglican) is orthodox theologically and is a social conservative. Yet he is on the Left (although he claims "never to have been a party man") politically. I think from what I can muster from his writings that he advocates some kind of decentralized distributivism ala G.K. Chesterton.DanaMcgee_{April 11, 2007
April
04
Apr
11
11
2007
03:52 PM
3
03
52
PM
PDT}

"I am an ID advocate and a socialist of sorts."
Hey Dana, me too. :)Atom_{April 11, 2007
April
04
Apr
11
11
2007
03:29 PM
3
03
29
PM
PDT}

By the way, what I meant by Communism was the Marxian brand. I am an ID advocate and a socialist of sorts.DanaMcgee_{April 11, 2007
April
04
Apr
11
11
2007
03:26 PM
3
03
26
PM
PDT}

I guess an even easier way to cut to what I am getting at is by asking the question: If Natural Selection refers simply to Differential Reproduction, what is Differential Reproduction the cause of? Be careful how you answer. If you say "greater complexity", that can be shown false; some replicators replicate more in a simplifed state. If you say "Differential Allele Frequencies" then that is just another phrase for "Differential Reproduction", on the level of genes. If you say "Greater Fitness", what is the general definition of "Fitness"?Atom_{April 11, 2007
April
04
Apr
11
11
2007
03:22 PM
3
03
22
PM
PDT}

"two" not "to" Also, if you say the variations themselves are what cause the effect in scenario B, then the causal aspect of NS refers to these variations (mutations) as the force. Since variation is random in Darwinism, then NS is a random cause. This is the point I am making. We also see that NS is no magic wand; organisms must do all the hard work themselves.Atom_{April 11, 2007
April
04
Apr
11
11
2007
03:04 PM
3
03
04
PM
PDT}

Prev 1 2 3 4 5 Next

You must be logged in to post a comment.

Leave a Reply