|
Author
|
Topic: Nature Refutes ID?: The Evolutionary Origin of Complex Features
|
Roger R
Member
Member # 200
|
posted 10. June 2003 07:35
RBH writes,
quote:
It's been asserted, but not specifically shown, that the evolved programs do not meet one or another aspect of Behe's or Dembski's definitions.
We might have a standoff here. I believe it was you, amongst others, who came into this thread proclaiming that the EQU program represented the evolution of IC. That would seem to put the ball in your court to establish a rigorous case for the assertion that it meets Behe's requirements. Not only have we not seen that, but we have seen you trying to rationalize discarding a significant part of Behe's definition, yet still claiming to meet the original definition. Either course might be productive, but both at once casts suspicion upon your enterprise.
quote:
1. single system. Do you argue that the program is not a "single system"?
Yes, I would argue that the program is not a single system in Behe's meaning. Indeed, wasn't it one of your points that the program is also responsible for reproducing? In the case of the BF (bacterial flagellum), we understand that the organism must also reproduce, in addition to other functions, but those aren't considered part of the IC core function. In addition, the programs are also capable of doing other logical operations, NAND by native instructions, others by assembling a series of instructions. Now, either those are part of the single system, and hence EQU isn't IC, since the system would be one of producing logical functions, and a knockout of EQU may indeed leave logical functions still operating, or they aren't, in which case the program isn't a single system.
quote:
2. several parts. There are 60 instructions in the Case Study program. Do you argue that there are not "several" parts in that program?
3. well-matched parts. Fuzzy, but do you argue that the several instructions aren't "well-matched" in the face of the fact that if they are arranged in a particular configuration, they perform logic operations? Do you argue that they are not "well-matched" even though the primitive operations that the individual parts perform mesh together to accomplish a complex logic function?
I would argue that the instructions are not "parts" in the Behe meaning, and that they are clearly not "well-matched" in relation to the EQU function. Indeed, there seems to be no relation to the knockout test and the logic of the EQU function as defined in the program. The knockout test used seems to relate merely to the logic of the program, and has little to do with the logic of EQU.
Just because they can be assembled together to produce some characteristic or property doesn't make them well-matched. Indeed, by a set of lego's and you can construct a lot of things, precisely because they are not "well-matched" to any one, but generically useful to assembling many. So although a knockout of a nop-a may or may cause the EQU function to be lost, that doesn't make it well-matched to the function.
quote:
4. interacting parts. Do you imagine that the various instructions don't interact, with the result of one instruction being an input to another?
5. parts contribute to the basic function. This is the focus of the operational definition. Recall that it was Behe who offered that operational definition, not me. Do you argue that the various instructions don't "contribute to the basic function" in the face of the fact that if we knock out any one of 35 instructions in the Case Study program it can no longer perform the logic operation EQU?
Putting aside the parts issue, I'll give you that they interact, but w/o the rest, you don't have a definition of IC. Many things interact which aren't IC. And some of the instructions contribute to the basic characteristic, although others do not yet might cause the EQU characteristic to disappear because they influence program flow.
Let me also say a word about "function". I hate to do this in light of the semantic difficulty we already seen in this thread, but I think it is important. Function can be used to describe EQU in accordance with logic / math terminology. But the use of the word function by Behe is a more generic use. There is a flaw introduced when to try to map the Behe expression "basic function" to a logical function such as EQU. It can have, by definition, no "basic function". It is a deterministic function. It either is EQU, or it isn't.
IP: Logged
|
|
YZ2
Member
Member # 91
|
posted 10. June 2003 09:33
Two questions have been raised, one related to testprime in nature, the other related to why testprime is relevant. These are my simple explanation.
1) Testprime can be done very easily if no claim of sufficiency. That is also why the mathematicians who found the sufficient solution are so celebrated, probably because they can think so clearly than many of us. So if evolutionists do not claim "sufficiency", ie. "not all ICs are evolved", there is no argument here, and we can all go home.
2) On why "TESTPRIME" is relevant here, is due to, again the argument of sufficiency. Indeed, it is not the definite choice, but a reasonable and perhaps a more convincing choice to show some evidence of sufficiency. You see, EQU cannot cut it, because intuitively EQU is hardly a convincing example of complexity, at least to most thinking people as a comparison to "ALL complexity of life". So in order to select a convincing representative example, I suggested testprime. It appears to be reasonable. It can be generated by "intelligence". It can also be generated blindly given infinite time and memory. So if evolutionists claim "ALL ICs are evolved." Why can't it be done perfectly with the TESTPRIME, now that we have the Avida environment to test it?
[ 10. June 2003, 09:53: Message edited by: YZ2 ]
IP: Logged
|
|
GP
Member
Member # 570
|
posted 10. June 2003 10:07
quote:
So if evolutionists do not claim "sufficiency", ie. "not all ICs are evolved", there is no argument here, and we can all go home.
If it were only that easy, YZ2. ![[Smile]](smile.gif)
There two rather obvious flaws in this analysis:
1) Sufficiency, as I had argued before, is empirically difficult to establish. And it seems from the discussion I've had with other ID proponents that there is no way to control for "front-loading." I don't see how one can claim anything more strongly than this -- Darwinian mechanisms are necessary for the evolution of IC structures, if they are evolved.
2) What "no ICs are evolved" implicitly demands is a truth statement. Once again, it does not seem reasonable that this statement can be established empirically. For every IC structure that a Darwinian tests and finds probable cause for evolution, another IDist comes up with another test. See where this is going? Now, c0uple this problem with the fact IDists are still struggling to come with a definition they are happy with (if there is any doubt, one should just look at Roger's post above, where he's juggling "well-matched" and "parts" and "interacting."). We might as well be testing the assertion "no green-eyed monsters are evolved."
This in my opinion is why your TESTPRIME example will not be received sympathetically. I do not see how the experiment would settle the negative argument that "no ICs are evolved." Maybe that's cause for people to go home anyways .
IP: Logged
|
|
GP
Member
Member # 570
|
posted 10. June 2003 10:14
quote:
You see, EQU cannot cut it, because intuitively EQU is hardly a convincing example of complexity, at least to most thinking people as a comparison to "ALL complexity of life". So in order to select a convincing representative example, I suggested testprime.
One more comment, before I sign off for today. Both you and Jack (and maybe Micah) have made a similar comment to the effect that "intuitively" and "not surprisingly" EQU did evolve. Though intuitions are sometimes helpful to guide research, they hardly serve as a convincing argument. Ex post facto analyses are rather subjective in their nature, and reveal personal biases. Tell us, YZ2, what prevents another IDist from coming along and claiming that TESTPRIME "intuitively" should have evovled given some initial conditions?
IP: Logged
|
|
Argon
Member
Member # 276
|
posted 10. June 2003 10:22
YZ2 writes:
quote:
You see, EQU cannot cut it, because intuitively EQU is hardly a convincing example of complexity, at least to most thinking people as a comparison to "ALL complexity of life".
I don't think anyone here suggested that the Avida simulation demonstrates how all the "complexity of life" evolved. Further, few scientists would claim that "all ICs evolved" (for example, the Acme mousetrap didn't), or that all possible IC systems are accessible to evolution within a given amount of time. What RBH and others are saying is that "ICness" might not be a reliable indicator of design.
See the difference?
[ 10. June 2003, 10:27: Message edited by: Argon ]
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 10. June 2003 11:17
Y2Z: So if evolutionists claim "ALL ICs are evolved."
Nope, what evolutionists may claim is that IC is not a reliable indicator of design.
IP: Logged
|
|
YZ2
Member
Member # 91
|
posted 10. June 2003 11:54
I do not quite follow the logic from some of the comments here.
What I said before are:
1) "Not all ICs are evolved", and not "no ICs are evolved".
2) "All ICs are evolved" should be "All ICs (in biology) are evolved".
3) To show negation of "All such-and-such", I use "Some such-and-such".
4) To show "sufficiency" here, I adopted a method to show "validity from a convincing representative example". In another word, to show "All is true", I use "a convincing representative example is true". There are flaws with this approach, but this is the one I came up with for now.
A genuine question for clarification:
From evolutionists' standpoint, it is "not ALL ICs are evolved"? What is the explanation for some ICs if they are not evolved, from an evolutionist's standpoint?
[ 10. June 2003, 15:33: Message edited by: YZ2 ]
IP: Logged
|
|
Jack Foster
Member
Member # 79
|
posted 10. June 2003 13:35
Hi all,
I find myself in sympathy with Roger's position. Behe not only provided a definition; he also provided an example: the mousetrap. The parts of a mousetrap interact directly with each other, and are well-matched for that interaction. If you view instructions as "parts", then it's also clear that they do not interact directly with each other; they interact indirectly through the intermediary CPU. If they do not interact directly, then they cannot be well-matched. Again we have Behe's example: the mousetrap. Behe could have used other examples, if he meant some other form of interaction.
Yet I still concede that the Lenski system has evolved IC. There absolutely must be some kind of virtual evolution analogue for IC. What is it? Is it coordination between subroutines that allows some function? . . . between lines of code?
I think the analogue must be the discovery through IC pathway (unselected steps) of some function that requires some degree of minimum (and irreducible!) complexity to accomplish. The paper reports that the performance of EQU requires a minimum of five nand operators, and that there were deleterious (unselected) mutations along the pathways to the evolution of EQU. That's enough to convince me.
I think ID looks pretty bad if its leadership claims that IC just doesn't exist in the virtual world.
RBH, now that you understand me, let's move back to evolvability for a second. (It still deserves it's own thread.) I think RBH imagines nature thru environment throwing different axes and fitness functions at an unevolvable replicator until the code is cracked and the unevolvable becomes evolvable. Evolvability is more properly understood as an intrinsic property of the replicator itself that exists regardless of external forces. External factors might determine how a replicator evolves, but they do not determine whether a replicator evolves.
regards,
[ 10. June 2003, 13:59: Message edited by: Jack Foster ]
IP: Logged
|
|
YZ2
Member
Member # 91
|
posted 10. June 2003 14:02
I am still quite puzzled, and perhaps surprised by this:
From the evolutionists' standpoint, is IC (in biology) evolved or not evolved?
IP: Logged
|
|
Argon
Member
Member # 276
|
posted 10. June 2003 17:06
YZ2 writes:
quote:
From the evolutionists' standpoint, is IC (in biology) evolved or not evolved?
In this post, I'm going to narrow "evolution" to mean "naturalistic biological evolution without the added, episodic intervention of an intelligent designer". For instance, Behe is an evolutionist, but not a supporter of entirely naturalistic evolution. On the other hand, Denton feels that the laws of nature were designed such that life would naturally evolve as a consequence. So he would support naturalistic evolution (at least within the parameters set by the creator of this universe).
I'm also going to restrict "IC" to the first iteration provided in Behe's DBB, which includes "knock-outs" as a test method (I'm not certain the criteria for the subsequent versions can be readily applied to most candidate systems).
The bottom line:
It is not known whether all biological IC systems evolved. However, if the working hypothesis is that all organisms evolved then it follows that IC biological systems would have too. There would be some haziness around the area of abiogenesis, however. For example, it is possible that the first cell did not evolve "naturally" (designed and deposited by terraforming aliens perhaps?) but that its subsequent progeny did. In that case, some IC systems would not evolved but a great deal would have. Personally, I have no idea how the first cells got started, and I'm not certain where eukaryotes, eubacteria & archaebacteria "root", and so I'm potentially open to ideas like alien terraforming and other mechanisms of intervention. Minimally, because I haven't seen much evidence for the frequent, episodic intervention by intelligences, I think a fair number of IC systems did evolve.
[ 10. June 2003, 17:21: Message edited by: Argon ]
IP: Logged
|
|
YZ2
Member
Member # 91
|
posted 11. June 2003 09:30
Thanks, Argon.
I should have used the term "ID-free evolutionary theory" previously instead. Perhaps, one can be an ID-evolutionist after all.
IP: Logged
|
|
Argon
Member
Member # 276
|
posted 11. June 2003 10:49
YZ2,
I think that the majority of ID-proponents that participate on this board are "evolutionists", meaning that they generally accept the idea of descent with modification.
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 24. June 2003 23:32
I wondered whether to start a new thread with this, but it seems to fit with the Lenski, et al work so I'll introduce it here.
I have just read one of the newer (2002) papers on Adrian Thompson's Hardware Evolution site, Notes on Design Through Artificial Evolution: Opportunities and Algorithms. quote:
'Opportunities' offers some remarks on the evolution of radically unconventional designs, and gives an example of an evolved nano-electronic design that employs dynamical principles previously seen in the literature of neuroscience, but not of nano-electronics. The `Algorithms' section then provides food for thought on what types of evolutionary algorithm may be able to tackle challenging design problems in the future, given developments in computer architecture. An example shows how a very simple mutation-driven algorithm can arrive at a surprisingly sophisticated design if neutral evolution is allowed.
The material on neutral mutations is particularly interesting. They might be comparable to the "stage-setting" mutations seen in the Lenski, et al paper. While I may revive this in the "Hardware Evolution" thread on Arn, it strikes me as relevant here, too.
Some of the hardware evolution studies have employed what amounts to a knockout procedure to evaluate the roles played by various components, which could provide indications of irreducible complexity having evolved there too. In at least some cases reported in Layzell's dissertation the knockout procedure on occasion abolished the evolved function, indicating that an irreducibly complex hardware device had evolved. I think this vitiates the incipient argument by Micah and Jack Foster that had reference to 3-D physical structures and Behe's mousetraps, since the circuits that evolved were instantiated in real physical hardware as they evolved, not as software simulations or as Lenski et al.'s assembly language programs performing logic functions.
The hardware evolution work is by and large indifferent to biological implications - those guys are primarily interested in design. In the paper above Thompson argues that there is a class of design problems for which evolutionary algorithms are not only indicated, they are indispensable.
RBH
[ 24. June 2003, 23:43: Message edited by: RBH ]
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 25. June 2003 13:10
RBH: An example shows how a very simple mutation-driven algorithm can arrive at a surprisingly sophisticated design if neutral evolution is allowed.
YES!!! This maps quite nicely on the findings about RNA and DNA namely that proteins and RNA are scale-free, and that their distribution is characterized by a few very common and many very uncommon structures distributed throughout sequence space and close neighbors to other structures.
This means that allowing for neutral evolution enables evolution to 'explore' throughout neutral space until a 'fortuitous' mutation brings it to a neighboring structure after which selective evolution can take over.
The structure of RNA and protein space already suggests that neutral evolution is an important factor.
Exciting. RBH, thanks for the links to these papers, I will be reading some of them in the next few days.
IP: Logged
|
|
Eric T. Malroy
Member
Member # 536
|
posted 03. July 2003 16:57
All,
Evolutionary algorithms are useless when the design space is excessively large, and systems are significantly coupled. Coupled systems result in irreducibly complex systems. Those who use evolutionary algorithms must artificially constrain the design space and de-couple the systems. This hidden logic enables the evolutionary mechanism to operate, but the modeling fails to simulate biological systems and the origin of complexity inherent to these systems.
Darwinists must explain how the design space in biological systems is constrained to enable the specification required to evolve new coupled systems. The typical gene has a thousand nucleotides linked together, where there are one of four nucleotides per position (A,T,C,G). This means that the design space is staggering- 4^1000 = 1.14813E602(considering other constraints it would be more closer to 3^1000 = 1.322070819E477). In addition, there are about 100,000 genes in complex life-forms. Given that the nucleotides are possibly coupled, the genes are coupled and the systems within the life-form are coupled, Darwinists are hard-pressed to explain the evolution of complexity in biological systems.
IP: Logged
|
|
|
Printer-friendly view of this topic
