|
Author
|
Topic: Response to Howard Van Till
|
kyle7
Member
Member # 191
|
posted 29. September 2002 07:21
Frances,
I have a problem with those who treat a hypothesis as if it were evidence. Many naturalists seem to like to pontificate about the unknown details of evolution. They seem to regard speculation as equivalent to evidence. The logic goes as follows:
1) Only naturalistic explanations are rational and scientific.
2) Other explanations besides the natural are unscientific and irrational.
3) If I can theorize about some supposed evolutionary pathway and come up with a possible explanation (no evidence), then my speculations can be added to the balance of evidence for evolution.
I find this thinking irrational and counter-productive for science. We need to hold on to our scientific theories with a degree of detachment lest we become philosophizers of science rather than true scientists.
One of your arguments that John (and you) mentioned was the problem of eliminating design. The main point about the car argument was to show that you also could never completely eliminate natural explanations. If ID arguments are invalid then naturalistic arguments are also invalid.
I also showed how the Second Law of Thermodynamics is a law that is stated as a negative – "it is impossible ... " This means that to completely prove the law we must examine all possible engines (an impossibility), yet scientist today accept the law. In other words, there exists two sets of engines. There are all the known engines (set A) and there are all the yet to be discovered engines (set B). Using your logic, we should not accept the Second Law of Thermodynamics and classify it under the "I don’t know" category given the existence of set B. You say the following:
quote:
I believe that Dembski, in his latest response to van Till, has identified a major short comming in the ID filter as proposed in the design inference and No free Lunch.
When van Till points out that one has to differentiate between an inference based on all natural cases versus known natural causes. Van Till denotes the former one with N and the latter one with n. Dembski argues that if our knowledge increases the probability may in fact decrease rather than increase. One may argue however that science has been quite exemplary in showing that in many cases the probability increased when a natural explanation was found for something which was initially attributed to design. History is full of examples. The question remains, may there be examples in which science will fail? Surely that possibility can not be eliminated but science unlike ID does not deal in proof by elimination.
So we need to accept that there will be instances in which our lack of knowledge may lead us to infer design where design is not the explanation when we gain a fuller understanding of the facts.
We also may find that our lack of knowledge is not going to be resolved by science and we may have to infer design. In such cases however should we infer design or chance? Why should one be privileged over the other? Yet Dembski seems to suggest that ID has to deal only in known processes, not yet to be discovered processes, but if I am correct then ID has opened itself up to false positives. Indeed the best way to avoid false positives is to allow "we don't know" as an explanation.
I want to highlight where you say that "science unlike ID does not deal in proof by elimination."
Clearly, the Second Law is a case where we are dealing with "proof by elimination". Consequently, your argument against the ID filter is shown to be false!
You end your last post by asking me, "perhaps you can show us an example in which Dembski's filter worked to detect design in biology?" The issue being addressed is if the design filter is unscientific given the existence of set B – this case being the unknown phenomena that may explain the specified complexity. We are talking about the validity of the method and not about if Dembski’s filter has detected design in biology. I would put forth the bacterial flagellum as a good argument for design, though the evidence or lack of evidence of design does not invalidate the design filter. The design filter is a tool to identify design. It will either establish evidence for design or it will not. ID theororists are just beginning to identify possible evidence for design. They have their work cut out for them.
IP: Logged
|
|
Elend
Member
Member # 326
|
posted 29. September 2002 08:00
kyle7:
In fact all the laws of physics can not be completely proven, not only the peculiar form of 2LT you put forward. One can never verify every instance when a law should apply. All the laws describe the highly likely behavior and are not hold to be absolute truths. And the 2LT is by no means a case where we are dealing with proof by elimination. But maybe it is not clear what some of us understand by "proof by elimination".
First we are interested in a set up that can be examined. (this is not the case for the Set B of engines you talk about). Say given the examined facts we could conclude A, B, or C, according to our present knowledge. Now, it turns out that none of these fits the facts. A proof by elimination would use this to conclude D. A scientific proof would not jump to concluding D, but gather more knowledge and form positive evidence for D. Examining the fact again it should point to D, BUT NOT BECAUSE A, B, or C failed, but because the evidence points to D!
In this light, 2LT is by no means "proof by elimination".
IP: Logged
|
|
Daniel Edington
Member
Member # 421
|
posted 01. October 2002 14:24
Dembski,
I have some minor issues with your rebuttal...
"Howard Van Till's review of my book No Free Lunch exemplifies perfectly why theistic evolution remains intelligent design's most implacable foe."
Is everyone who disagrees with your particular views considered your enemy?
"But the genes follow the proteins which follow the function, and not vice versa, so my analysis is the correct one."
I am not so sure you have this one correct, I think it should be more like proteins follow genes and the function follows the protein. Unless you can justify that statement, I think you might want to reconsider your analysis.
Dan
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 01. October 2002 22:08
Kyle:
I believe that you are wrong about the 2nd law of thermodynamics being eliminative. To show this I can make a positive claim for the SLOT namely that the entropy of a closed system can only increase. Can you make a similar claim for ID? I doubt it because ID is infered through true elimination.
IP: Logged
|
|
Paul A. Nelson
Member
Member # 26
|
posted 02. October 2002 09:11
Daniel Edington wrote:
quote:
I am not so sure you have this one correct, I think it should be more like proteins follow genes and the function follows the protein.
Why does the function follow the protein? And why does the protein follow the nucleic acid?
(Dan -- just to forestall the obvious response: you can assume I'm familiar with the Central Dogma, etc.)
[ 02. October 2002, 11:50: Message edited by: Paul A. Nelson ]
IP: Logged
|
|
Mark Szlazak
Member
Member # 391
|
posted 02. October 2002 10:40
Hello Frances,
This may sound naive but I've got a question about the 2nd law of thermodynamics. Which version of the 2nd law are you using in your arguments and could you state that version of this "law" in detail?
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 02. October 2002 10:49
Mark:
2LOT: Entropy can never decrease in a closed system
2LOT: Entropy always increases in a closed system
Kyle was arguing that the 2LOT was eliminative but I guess Kyle confused eliminative with a 'negative' formulation: "Entropy never decreases in a closed system".
Kyle used the following definition:
"It is impossible for a cyclically operating device to convert thermal energy to mechanical energy without wasting some thermal energy to the cold surroundings."
thus a 'positive' version could be:
All cyclical operating devices waste some thermal energy when convering thermal energy to mechanical energy
All in all it seems to be clear that the SLOT is not eliminative like for instance the Design Inference ala Dembski.
IP: Logged
|
|
Daniel Edington
Member
Member # 421
|
posted 02. October 2002 15:15
Paul,
"Why does the function follow the protein? And why does the protein follow the nucleic acid?
(Dan -- just to forestall the obvious response: you can assume I'm familiar with the Central Dogma, etc.)"
Why not? My rational is based on the fact that proteins have function because of thier structure and that structure follows from the amino acid sequence. The amino acid sequence follows from the genome.
So anyway, what is this central dogma? What is the obvious response?
Dan
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 02. October 2002 17:23
quote:
Why not? My rational is based on the fact that proteins have function because of thier structure and that structure follows from the amino acid sequence. The amino acid sequence follows from the genome.
So anyway, what is this central dogma? What is the obvious response?
Uh...
http://www.accessexcellence.org/AB/GG/central.html
You mentioned it yourself, perhaps you were just unfamiliar with the term? Anyway, it is quite an odd term to use in science, I guess we have to blame Crick or somebody.
IP: Logged
|
|
Daniel Edington
Member
Member # 421
|
posted 02. October 2002 22:49
yersinia,
Thanks for clearing that up. Because I am not a molecular biologist I was unfamilar with this terminology (Nice picture to.)
I had a particular problem with this statement:
"But the genes follow the proteins..."
There is no basis for this statement. Is not the implication of the above statement that if a foreign protein is introduced into a cell, the DNA sequence to code for that protein should be produced? We know this does not happen, but when foreign DNA is introduced into a cell the corresponding protein can be formed.
So in reality when computing the probability of the bacterial flagellum forming, shouldn't we be attempting to determine what the probability of the generation of the correct DNA sequence that code for the system in question?
Just a thought...
Dan
IP: Logged
|
|
Paul A. Nelson
Member
Member # 26
|
posted 03. October 2002 09:32
Thanks for the nice pic, Nic (once again you show yourself the Master of the URLs!).
Dan, you wrote:
quote:
My rational is based on the fact that proteins have function because of thier structure and that structure follows from the amino acid sequence. The amino acid sequence follows from the genome.
Yes -- that's the Central Dogma in a nutshell.
But notice in the diagram above, one has a protein emerging at the bottom, ready to fold into its functional conformation. Thus a further arrow could be added, from the protein to some cellular role, indicating "function" -- the raison d'etre of the whole story. Why do cells bother to have genes encoding proteins? Because they (cells) have functions to perform.
I think you can see Dembski's point if you start with a "Why?" question. The terminus of that line of questioning will not be any particular molecule (e.g., DNA), but rather a biological function.
IP: Logged
|
|
Daniel Edington
Member
Member # 421
|
posted 03. October 2002 14:21
Paul,
I really can't see his point. Your statement that DNA encodes for protein because the protein has some function to perform has no real empirical basis, it is an assumption based on your predisposition toward intelligent design. It may be correct, or it may not be, but we can’t really know this. The only thing we know for sure is that information flows from DNA through RNA to the final heir of this information, a protein. The actual function of the protein may in fact be incidental. We could do an experiment using recombinant DNA technology to insert a sequence of DNA into a bacteria, that codes for a more or less random sequence of amino acids. The bacteria’s cellular machinery would synthesis this protein just like it would any other, even if it has no biological function. The fact is that an RNA polymerase really doesn’t care if the mRNA that it is working on has a biological function or not.
Irregardless of what point Dembski was attempting to make, the fact is that no matter how a new functional systems arose in a cell (be it by intelligent design or some natural mechanism) it seems logical that it was the genetic sequence for such a system that came about first and the proteins followed after.
Dan
IP: Logged
|
|
Janitor@MIT
Member
Member # 125
|
posted 03. October 2002 16:17
I understood that “information” flowed the opposite direction. From natural selection through function to DNA. Since there is a circle of causality here, NS > DNA > NS >..., I don’t know why we can’t, as is often done in evolutionary biology, reverse the usual “logic” of the train. DNA <> NS. I mean, it works either way doesn’t it? I.e., if we ignore everything that comes in between…
IP: Logged
|
|
RB
Member
Member # 263
|
posted 03. October 2002 21:15
Paul Nelson writes: I think you can see Dembski's point if you start with a "Why?" question. The terminus of that line of questioning will not be any particular molecule (e.g., DNA), but rather a biological function.
Which is why science does not ask such vague WHY questions, but more specific why questions, usually noted by the word HOW.
Why is the sky blue vs How is the sky blue.
I know this is way to simplistic and a philosopher of science,like yourself, could tear it up, but ultimately the best science moves (even when through some sort of revelatory AHA moment) through very specific questions.
ie. Why do flies use bicoid as an anterior morphogen vs how does bicoid work as an anterior morphogen.
One could ask why do flies even bother to have bicoid, when the rest of the insect world can get by without it. That would lead to another interesting question, is bicoid really an example of a biological specification, if there are multiple ways to specified anterior/posterior axes in insects. (new sub-question, how do grasshoppers and beetles make heads without bicoid) A whole lot of interesting questions.
IP: Logged
|
|
Paul A. Nelson
Member
Member # 26
|
posted 04. October 2002 12:06
Dan wrote:
quote:
Your statement that DNA encodes for protein because the protein has some function to perform has no real empirical basis, it is an assumption based on your predisposition toward intelligent design.
That will come as a surprise to the sequence annotators at GenBank, who classify DNA sequences by the functions they encode: structural protein, transcription protein, cell division protein, etc. One hardly needs to be predisposed to ID to recognize the primary of biological function in understanding DNA.
Dan wrote:
quote:
The only thing we know for sure is that information flows from DNA through RNA to the final heir of this information, a protein.
And how does the information get into the DNA? Janitor@MIT suggested the most common answer: "From natural selection through function to DNA." The process of selection preserves function. Even from a Darwinian standpoint, therefore, function is causally primary. This would be true even in the most stripped-down RNA World scenario (ask yourself why one replicator comes to dominate a population of catalytic RNAs).
Dan wrote:
quote:
We could do an experiment using recombinant DNA technology to insert a sequence of DNA into a bacteria, that codes for a more or less random sequence of amino acids. The bacteria's cellular machinery would synthesis this protein just like it would any other, even if it has no biological function.
Do you think that is a plausible model for the origin of the information in DNA, RNA, or protein?
What you're describing is simply information transfer through cellular machinery: DNA makes RNA makes protein (i.e., the Central Dogma). But the Central Dogma (as expressed in the machinery) does not explain itself, any more than one can generate a photocopier by using that device to reproduce ink blots over and over.
What we know from biology tells us that DNA, RNA, and protein co-exist, and are interdependent -- and the thread common to all, and that all presuppose, is biological function. You write that it seems "logical" to start with DNA, but nucleic acid that codes for nothing (random strings of amino acids) will never give one a single protein, much less a living cell. Anyone actually constructing a cell will need to ensure that the complete network of functional relationships - function -- DNA -- RNA -- protein -- function - is in place. DNA on its own is an inert molecule.
In short, to have real biology, one needs functions. That's the case from either an evolutionary, or design, standpoint.
RB wrote:
quote:
Which is why science does not ask such vague WHY questions, but more specific why questions, usually noted by the word HOW.
Nope: "how?" won't do it. I meant ordinary functional "why" questions. Consider a non-biological example.
Q. Why does my laptop have a fan and vents?
A. To dissipate heat from the CPU, motherboard, and power supply.
Q. Why does heat need to be dissipated from those structures?
A. Because they won't operate properly above certain temperatures.
And so on. These kinds of "why" questions, within what is often called reverse engineering, figure centrally in biological reasoning. As John Maynard Smith (no IDer!) writes,
quote:
Biologists often find themselves confronted by a structure and ask themselves what function it was selected to perform. Harvey's discovery that the heart is a pump was an early triumph of such reverse engineering.
Before he became an evolutionary biologist, Maynard Smith worked as an aircraft designer:
quote:
In 1945, when I was working in an aircraft design office, the Royal Aircraft Establishment at Farnborough put on an exhibition of recently captured German equipment. A friend and I spent two fascinating days looking at the equipment and asking ourselves, "Why did they do that?"...I seem to remember that the V2 rocket had a gyroscope puzzlingly connected to the fuel supply to the motors; surely, one would think, it should be connected to the guidance system....Since I became a biologist, I have spent most of my time asking questions like that. (J. Maynard Smith, New York Review of Books, 30 Nov. 1995, p. 46)
"How?" (unless it conceals an implicit "Why?") won't work for functional questions, which concern ends-means relationships.
[ 04. October 2002, 12:22: Message edited by: Paul A. Nelson ]
IP: Logged
|
|
|
Topic Closed
Printer-friendly view of this topic
