|
Author
|
Topic: Stone circles
|
yersinia
Member
Member # 324
|
posted 25. January 2003 10:05
Cre8tionist,
The whole point with the stone circles example was that the circles are not predictable based on the physics and chemistry of ice and rock. Neither one "crystallizes" into circular shapes, and certain not at that scale.
A complex self-organization model was developed and published in the Science paper. It ends up being a complex, repetitive *interaction* of "law" + "chance" that produces these circles.
Unfortunately Demski does not handle the law+chance possibility very well in his various publications; he usually just lumps them into the chance category, even though the combination (e.g. stone circles...or evolution for that matter) can do things unimaginable on a strict chance/law dichotomy...
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 25. January 2003 12:45
The stone circles provide another natural phenomenon on which to test and validate the design detection methodology proposed by Dembski. In the Validating Design Discrimination Methodologies thread I wrote quote:
Somehow or other the various design discrimination methodologies, whether they are the 'design themes detection' of MDT or the meat axe design/no-design EF discrimination of SDDID, are going to have to be validated and calibrated, and human designs constitute one universe in which to do that. There are others, of course. Evan suggested that using biological phenomena where we are confident that design is not occurring as a test bed for development. That's a good suggestion. Brauer, on the Solar eclipses thread, suggested a physical phenomenon, and I developed that in a discussion of the Oklo reactor. There are a multitude of other physical phenomena where one can do precise calibrations of a design discriminator. So there are at least three suggestions for different universes of phenomena in which testing and development of design discrimination methodologies can occur. That's all to the good, since wider validation testing implies more confidence in the outcomes of research on unknown phenomena.
Along with solar eclipses and the Oklo reactors, the stone circles are another test bed for evaluatiing and validating design detection methods.
However, Paul Nelson says quote:
I don't know how one would submit stone circles to the explanatory filter in what Andy calls a "naive" way -- which I take to mean something like, "Let's pretend we don't have a physical mechanism to explain these patterns." Nor can we reasonably answer the question Frances poses about how the filter would have worked in the 19th century. A critic would say the filter would fail; an advocate, that it would work.
If that's the case, then the explanatory filter is not the objective design detection methodology it purports to be. It is clearly a knowledge-dependent sorting device and thus is no more reliable than the idiosyncratic classifications of an unaided eyeball. One cannot do research using the EF in a researcher-independent manner if the classification that results from its application depends on what the person doing the classifying knows, and thus the EF is no advance over simply looking at a phenomenon and saying, "I can't think of a way that could occur naturally, so it looks designed to me!"
RBH
IP: Logged
|
|
Paul A. Nelson
Member
Member # 26
|
posted 28. January 2003 12:58
Hi all,
Back from Florida. (On Friday morning, I rescued a nearly frozen tiger-stripped anole from the sidewalk in downtown West Palm Beach; the poor guy apparently fell from a nearby palm tree. Took him to Scott Minnich's hotel room, where Scott warmed him up in the sink -- after a few minutes, he was running around as if an unexpected blast of cold Canadian air hadn't hit the area.) Anyway: several of the comments in the thread above leave me puzzled.
For instance: can anyone point to any scientific inference that does not depend on the "state of our scientific knowledge?"
IP: Logged
|
|
Irving
Member
Member # 535
|
posted 28. January 2003 13:25
Yersinia,
The ariticle mentioned that Kessler & Warner developed a mathematical model that "explains how Mother Nature builds such bizarre shapes in the Arctic soils. " I'm curious as to how they modeled "chance" and their use of a randomizer. Any insight?
Thanks
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 28. January 2003 14:21
Paul A Nelson said:
quote:
For instance: can anyone point to any scientific inference that does not depend on the "state of our scientific knowledge?"
But I think that an important aspect is which of those scientific inferences of the past becomes absolutely invalidated and completely reversed by a future refinements of knowledge.
Examples to consider:
Even “music of the spheres,” which had a philosophical component, did provide something like a series approximation to the motion of the planets. As such, it is still valid as a calculation, though of relatively low accuracy. Certainly overturned but only to a matter of degree. Gravity still causes motions in something of a spherical (actually nearly conic section) pathway.
Newton’s laws, are strictly overturned now. But classical mechanics still serves about as well, possibly even better, for modeling a great deal of celestial motion problems, including rocket launches. Only the tiniest of perturbations need to be dealt with in some unusual circumstances.
So indeed, these elements of scientific inference depend on the “state of our scientific knowledge.” Which of these scientific inferences would be completely overturned by finding a new physical mechanism?
But ID inference, since it is in the form of an “argument from ignorance”, would be so overturned. (Of course I don’t consider it to be a “scientific” inference, because there are many difficulties of basic logic and ability to test by physical observational principles -- but that is another discussion.)
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 28. January 2003 15:13
What happened? Three (four?) posts put up this morning in response to Paul's latest short "Back from Florida" posting have disappeared. Is that a result of the changeover?
RBH
[ 28. January 2003, 15:14: Message edited by: RBH ]
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 28. January 2003 16:17
Yes, a couple of posts got lost. Of course, I didn't save mine, but the jest was that the whole point of an experimental control is to "ignore" previous knowledge to test a methodology.
I made the example of experimental controls during molecular biology assays. For instance, as John Bracht now knows well (or should ![[Wink]](wink.gif) ), when performing a PCR experiment it is almost imperative to run, in parallel to the experimental samples, a reaction tube in which no "target" DNA is introduced, and a tube containing a known positive sample. These are run just like the experimental samples, "as if" we did not know what we ourselves have put inside each tube, thus providing an internal validation to the experimental protocol.
This is standard procedure in science, and indeed every new technique must be tested repeatedly against known positive and negative controls to establish its sensitivity and accuracy.
Running the stone circles through the EF, "as if" we did not know what the underlying mechanism was, is not, as Paul suggested, an attempt to generate a scientific inference (of course that would be unwarranted, since we already know the answer), but an attempt to validate the EF against a proper, naturally originated (and hence more similar to what we may encounter during a real "experiment") negative control sample. I particularly like the stone circles because the mechanism for their formation is the same combination of chance and law that would be predicted to be able to generate design-like complexity by naturalistic means.
It truly is a rare chance, and I don't understand the reluctance to perform the test.
IP: Logged
|
|
Paul A. Nelson
Member
Member # 26
|
posted 28. January 2003 16:56
Guys,
![[Frown]](frown.gif) I responded to several posts here (including the original version of Charlie's), and my response, and the other posts, is gone (I hope temporarily). I'm going to wait a few hours to see if my response reappears -- it did in another thread (on ontogenetic depth) where the same thing happened.
P.S. Put another blue sad face here. Now the ontogenetic depth thread is gone too. Like an IDiot, I didn't save my lengthy reply in this thread, but I'll reconstruct it later today if it doesn't reappear. Ah technology -- live by it, die by it.
[ 28. January 2003, 17:05: Message edited by: Paul A. Nelson ]
IP: Logged
|
|
John Bracht
Member
Member # 5
|
posted 28. January 2003 18:31
Gedankin writes,
quote:
So indeed, these elements of scientific inference depend on the “state of our scientific knowledge.” Which of these scientific inferences would be completely overturned by finding a new physical mechanism?
And he goes on to say that ID has the problem that it can be overturned by finding a new physical mechanism.
Like Paul Nelson, I confess to finding this a rather strange argument. After all, even an evolutionary explanation could be overturned by finding a new mechanism! Imagine finding that the sequence of all known genetic information is found to be pre-coded into some non-biotic structure, and it can be shown that a law-like mechanism governs the formation of DNA from that non-biotic substrate and insertion of that "new" information into organism's genomes.
It's true that any theory can be overturned by new knowledge and new mechanisms--and that applies just as well to Darwinian theories as to ID theories. Darwinian theory claims to provide a mechanism to account for all of biological complexity. ID claims to overturn that mechanism by showing instances of biological complexity that fall outside the capabiilities of that mechanism (and yet these structures are within the known capabilities of intelligent agency); indeed, this represents an advance in our state of knowledge and not an "argument from ignorance". In reality, there are two mechanisms: Darwinian and design (yes, design is an observable mechanism operating in the world around us) and it is a matter of determining which one best fits the data. Obviously, new scientific knowledge could overthrow either inference--but that doesn't de-legitimize the practice of coming up with mechanisms to explain phenomena!
Imagine that I complained that evolutionary explanations were illegitimate simply because "new knowledge might overthrow them!" We could always gain further knowledge that would overthrow any given evolutionary explanation, or any design explanation. So when I hear this argument being propounded, I have to wonder whether those making it have really thought about what they are saying. The complaint is, itself, absurd.
John
[ 28. January 2003, 18:35: Message edited by: John Bracht ]
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 28. January 2003 20:21
But John, the problem here is not whether or not further knowledge may overthrow a design inference; I think even Dembski has recently stated that design inferences are provisional, and dropped (or clarified) his "no false positives" claim.
The issue here is whether the method currently proposed by (most) ID theorists to be the best available, and at the very least a "reliable" detector of design, the EF, is indeed so, or whether it has, as some of us suspect, some significant drawbacks that may result in a high frequency of false positives (among other things). To establish the accuracy and sensitivity of the method, a series of controls must be run through the EF, and the outcome evaluated. Otherwise, the EF will never be a systematic tool, but just an anectodal reference point with no practical utility whatsoever.
The stone circles just arguably represent an excellent control sample (there are more that have been suggested, some of which are also good, IMO). At the very least, if the EF is not applicable to problems such as stone circles or natural nuclear reactors, an argument must be put forth to explain why, in order to understand its limitations. This seems like basic methodological carefulness to me.
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 29. January 2003 01:22
But John,
quote:
Like Paul Nelson, I confess to finding this a rather strange argument. After all, even an evolutionary explanation could be overturned by finding a new mechanism! Imagine finding that the sequence of all known genetic information is found to be pre-coded into some non-biotic structure, and it can be shown that a law-like mechanism governs the formation of DNA from that non-biotic substrate and insertion of that "new" information into organism's genomes.
This makes my point! (Let me expand on how.)
The assumption here is that this “law like” mechanism governs what is “new” -- which implies that much stays the same. This implies descent with modification -- simply changing the mechanism of the “modification”. So in the first place, it would not overturn common descent. Therefore it would not overturn “Darwinism”. It would be vaguely like quantum mechanics overturning Newtonian dynamics, but not really because of other differences I will expand on later.
Now you said “…all known genetic information is found to be pre-coded into some non-biotic structure,…”. We know from examination of the genome that each person’s genetic makeup is distinct (though very similar) -- and this is pretty much true for broad ranges of organisms. So your statement does not only imply that some generalization is stored, rather that each individual’s genetic code (or at least a compressed form, as in a difference report) must be stored in that “structure”. And since you described it as a “structure” it must reside in some location. Thus the “insertion” process must communicate over worldly distances, do so with each newborn, and do so without prior detection.
I hope you see that such a new finding must also overturn much we know of physics. So in my opinion, this is about as likely to happen as finding that Newtonian physics was just a facade, that the sky really is a giant painting. Sure it could happen. Do you really propose such overturning as reasonable basis for your pont?
Then it’s not like QM overturning Newtonian physics, because QM (and relativity) must have a “correspondence principle” to explain why all those tremendous number of observations are consistent with Newtonian dynamics. We would, in John’s new “mechanism” still need to have all the genetic variations and individual to individual variations that we see now. We might even discover that the new “mechanism” was running a Darwinian algorithm to generate these new DNA sequences and find the ones that were fit for the environment -- and that was obviously likely because it provided a “correspondence principle” for the large number of consistent observations.
[ 29. January 2003, 01:33: Message edited by: gedanken ]
IP: Logged
|
|
Cre8ionist
Member
Member # 140
|
posted 29. January 2003 08:45
Found a few minutes and wanted to add one more comment:
In "The Creation Hypothesis, Information & the Origin of Life" Thaxton & Bradley showed that there is random/non complexity, ordered complexity and specified complexity. In so doing, they showed that things like the circles in question are ordered complexity. The following picture is generated from a program I wrote years ago demonstrating their point:
In their article they explain the relationship of these types of structures to information (they used letters, I used digits):
quote:
Information in this context means the precise determination, or specification, of a sequence of letters. We said above that a message represents "specified complexity." We are now able to understand what specified means. The more highly specified a thing is, the fewer choices there are about fulfilling each instruction.
In a random situation, options are unlimited and each option is equally probable. In generating a list of random letters, for instance, there are no constraints on the choice of letters at each step. The letters are unspecified. On the other hand, an ordered structure like our book full of "I love you" is highly specified. It has a low information content, as noted before, because the instructions needed to specify it are few. Ordered structures and random structures are similar in that both have a low information content. They differ in that ordered structures are highly specified and random structures are unspecified.
The naturally occurring circles are like the book full of "I love you," specified but low in information. Therefore not nearly as improbable as specified complexity. In the program noted above it takes more specified information to code for the digits on the right than either of the other two. In the random case, the code is choose red , then any digit between 0-9 and then do it again for x number of iterations. In the ordered structure, choose red , then any digit between 0-9 and repeat that same digit for x number of iterations, and in the specified case, choose specific digits and colors at varying steps of the process. So while it may appear that the circles being discussed in this thread are potentially past the upper limit of Dembski's explanatory filter, they are very low information structures and therefore much more probable than "specified complexity" structures or than may appear at first glance....................Cre8
[ 29. January 2003, 08:54: Message edited by: Cre8ionist ]
IP: Logged
|
|
Paul A. Nelson
Member
Member # 26
|
posted 29. January 2003 09:45
OK -- this time I saved my reply. Here it is again, lightly edited:
Guys,
Read the OP of this thread again. I’ve known about stone polygons in arctic regions, as the products of mechanical weathering action, since 1984 –- not since two weeks ago. The first step in the explanatory filter is to ask whether a natural cause explains an event. Physical mechanisms to explain patterned ground formation exist in the geological literature. The recent publication in Science is simply an attempt to explain “the range of forms...in a single model.”
So any realistic application of the filter would take one straight into the geological literature -– i.e., “to the state of our scientific knowledge,” which was already well-developed before the recent Science publication.
Thus, when Charlie writes,
quote:
The only way to know whether the EF is worth at least the paper it was printed on, is to test it precisely against things we know were designed, or weren't, while pretending we don't. Two weeks ago, using the EF on the stone circles would have been an "experiment" for the EF, allowing, presumably, some scientific inference regarding their origin. Today, it's an almost perfect control for its accuracy.
I can only scratch my head. This seems rather like saying, “Calculate the mass of the electron, but pretend you don’t know any physics.” The EF is constructed to lead one back into what we already know before moving on to design as a possibility. If someone applied the filter to stone circles, but didn’t (as his first analytical step) consult geological knowledge to see what had already been learned –- i.e., making sure that he’d surveyed scientific knowledge at the first (natural cause) node –- that would be a misapplication of the filter.
Here’s what I wrote a few years ago in my Mere Creation chapter, about this general point:
quote:
Design theorists do have a genuine responsibility, set by their own method, to consider natural causes. They discharge that responsibility, however, when they survey thoroughly what is known about natural causes: that is, when they consider what is empirically the case, as ordinary reasoners must.
I strongly favor testing the EF, but such any such test should not presuppose a misapplication of the EF. Of course Charlie is right: We shouldn't know, in any test, which patterns or events are naturally caused, and which intelligently caused (see my description of the proposed bit-string experiment, above); but we should be able to use all our scientific knowledge in the sifting (analytical) process. That's what we shouldn't pretend not to know.
In short, if we correctly applied the explanatory filter to arctic stone circles, our analytical path would take us straight back into existing geological knowledge, including the recent Science paper. Not a very interesting, or significant, test. Why not apply the filter, and other design detection methods, to data where only the sample compilers know which are the naturally, versus intelligently, caused patterns? This would not be a hard experiment to perform, and if a scientific or skeptics organization such as CSICOP agreed to assemble the bit strings and keep the key secure, the results would be significant indeed.
Anyone with ideas about how best to construct the bit-string experiment, please e-mail me at pnelson2@ix.netcom.com. Particular problems I'm trying to solve, in the experimental design, include how to define "naturally caused" and "intelligently caused," and also the issue of the security of the key.
Andy wrote:
quote:
I think the point is, Paul, that most of science methodology is aware of the provisional nature of the inferences. To my eyes, in contrast, Dembski's "explanatory filter" sets great store by its definitive nature.
Bill has dealt with this question in an on-line essay:
http://www.designinference.com/documents/2002.10.logicalunderpinningsofID.pdf
See the section on the reliability and assertibility of design inferences (p. 13ff.).
[ 29. January 2003, 10:46: Message edited by: Paul A. Nelson ]
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 29. January 2003 10:48
Paul A. Nelson said:
quote:
I strongly favor testing the EF, but such any such test should not presuppose a misapplication of the EF.
So I assume that in the proposed tests, that a finding of a failure of the EF to be “reliable” because of something initially missed in the checks for a not-so-low probability due to physical cause should not be counted against the EF’s “reliability”.
Go back to my first post in this thread. The problem with the EF occurs because it makes claims when the probability that all known pathways to an event add up to less than a “Universal Probability Bound” (UPB) of something like 10^-150. But what is the probability that a mistake was made in the analysis of these events, including having left something out of the analysis?
The result of the EF application can be overturned if your analysis has a greater than 10^-150 chance of being in error. The statistical tests required in scientific analysis would not normally allow such an error to have such a supersensitive effect on the result of the scientific analysis.
If you got up in court, and presented a case: This was person X’s DNA, and the probability that it was not person X’s DNA is less than 1 in 10 milliion.
Then the prosecution gets to ask questions: But Mr. Researcher, isn’t it true that if you didn’t consider some other explanation and the probability that you missed that other explanation is greater than 10^-150, then your analysis is totally incorrect? [Now at this point I am presupposing something that is not true -- that the correctness of DNA analysis would depend on your not missing something by the stupendously low probability of 10^-150] So you have to answer -- yes, if there was a chance of 1 in 10^150 that I made a mistake, this could be entirely wrong.
So how should the court rule on DNA evidence? [If that were a characteristic of DNA testing and errors in the procedure?]
But the fact is that there would have to be a 1 in 10 million or better chance that the researcher mixed up the tubes or something before the analysis of DNA became suspect.
However, if there is greater than 1 in 10^150 chance that the person claiming a successful application of the EF left something out of the analysis, then the EF should be overturned!
Given that, how can you ask that possibilities of mistakes in the analysis not be included in tests of whether the EF is “reliable”? The possibilities that tubes were mixed up in DNA analysis should be considered when offering evidence in court for the accuracy of a DNA claim, should it not? (I’m not claiming that is done, I’m just giving a case for this.)
quote:
Why not apply the filter, and other design detection methods, to data where only the sample compilers know which are the naturally, versus intelligently, caused patterns? This would not be a hard experiment to perform, and if a scientific or skeptics organization such as CSICOP agreed to assemble the bit strings and keep the key secure, the results would be significant indeed.
The fact is that this puts an inherent bias into the testing of the EF. Since all the “designed” patterns in the test sample will turn out to have a relatively high prior probability of an intelligent agent acting, these will mistakenly tend to validate the EF. The tests must include the case in which the ID advocates are willing to acknowledge no ID is active, and yet there is to those running the test no known knowledge of how a physical cause could exist for the event. The test must include the possibility of mistaken claim of no known physical mechanism to be a realistic test.
[ 29. January 2003, 12:32: Message edited by: gedanken ]
IP: Logged
|
|
andyg
Member
Member # 415
|
posted 29. January 2003 19:49
In both the article by Dembski that Paul Nelson referenced, and also in the post by John Bracht, we find a confusion between the mechanism of evolution and the details of a particular evolutionary scenario.
Bracht writes:
quote:
Darwinian theory claims to provide a mechanism to account for all of biological complexity. ID claims to overturn that mechanism by showing instances of biological complexity that fall outside the capabiilities of that mechanism (and yet these structures are within the known capabilities of intelligent agency); indeed, this represents an advance in our state of knowledge and not an "argument from ignorance".
In his article, Dembski writes:
quote:
The only way actually to demonstrate this [ the superfluous nature of ID] is to exhibit material mechanisms that account for the various forms of biological complexity out there........... But that hasn't happened. Why not? The reason is that there are plenty of complex biological systems for which no biologist has a clue how they emerged
Evolution, like design, is a mechanism to explain complexity. We know that things evolve, and we know a great deal about the mechanism by which - in principle - this happens. By this, I'm referring to variations in alleles in a population caused by, for example, mutation, and the changes in allele frequency in that population due to, for example, selection or genetic drift. I hope that Nelson, Dembski nor Bracht will dispute this.
We don't however, have any detailed descriptions of how this happens, largely because of the time periods involved and the difficulty in monitoring the genetic changes that occur over time.
Let's take a very simple example. The tetra fish Astyanax mexicanus exists in two forms - a surface-dwelling form, and a cave-dwelling form that is blind and has degenerate eyes
Cavefish web page
The two forms are completely interfertile, and thus qualify as members of the same species. In fact, there are over twenty different blind forms of mexicanus, all of which are interfertile, and fall into different complementation groups. This implies the existence of a series of different mutations, all of which led to the degeneration of the eye in one way or another.
We do not have a detailed description for how the surface-dwelling form of mexicanus turned into the blind form. We do, however, have a mechanism - evolution - and may soon even have candidate genes to describe how this change occurred.
All fine so far. Where Dembski and Bracht fail, I think, is that they assert that certain complex biological structures cannot be formed by the mechanism of evolution, without explaining why. Then, in a sleight of hand, Dembski implies that since there are many cases where we cannot provide a detailed scenario for how evolution of, say, the flagellum happened, then the mechanism itself is at fault.
This is no different from a critic of ID claiming that ID is inadequate because the nature of the designer is not known, or that an exact scenario for the design event is not given. In both cases, the argument is flawed. The accusation of argument from ignorance, and an appeal to a Designer-of-the-gaps still stands.
AndyG
IP: Logged
|
|
|
Printer-friendly view of this topic