|
Author
|
Topic: Exhaustive Sweep of Chance Hypotheses
|
Evan
Member
Member # 164
|
posted 16. June 2002 11:35
Hello Paul.
I agree with several things you said, but some ways in which I disagree with you are obviously not clear from my previous post. I would like to clarify some distinctions.
You write, “Figuring out that a monolith ... was intelligently caused is not analogous to a design inference. It is a design inference.”
I agree with this, and did not mean to say at all that a design inference in archeology, forensic science, etc. was in principle logically different than, much less merely analogous to, a design inference in biology.
What I said was that the referent of the design inference was merely analogous: the fact that biological objects are like, in some ways, designed objects with which we are familiar (such as machines made by humans) does not, in and of itself, constitute evidence for concluding that biological objects are designed. The analogies may “illuminate” our thinking about biology, but such illumination is not evidence.
You allude to this distinction when you write, “The point of analyzing design inferences in general, or as a type of inference, is to see what we can learn about the logical and evidential structure of such arguments.” (my emphasis)
The logical structure of a design inference is different than the evidential structure. In both the case of the artifact from archaeology, forensic science, etc. and the biological object, the logical structure of the design inference (based on the explanatory filter and considerations of probability) are the same.
However, the evidence that can be marshalled in support of the design inference is different.
In the artifact case, we know (directly, not inferentially) that intelligent agents exist (human beings) who have the resources (abilities, motivations, proclivities, etc.) to be considered as a possible cause. Therefore we can compare the artifact in question with both natural objects and other objects that we know have been designed by humans - our suspicion that something is “not natural” is conditioned both by our knowledge of nature and our contrasting knowledge of what known intelligent agents (human beings) can do. Thus we can marshal both “negative” evidence for the role of natural causes and “positive” evidence for the activity of a designer.
In the biological case, we do not have such “positive” evidence. We have no independent knowledge of an entity who can create biological structures. Therefore, only “negative” evidence against the conclusion of natural causes is possible.
This seems to be exactly what Dembski describes in his explanatory filter and in his arguments concerning CSI - that the existence of designed complexity is shown by establishing improbability in respect to natural causation - period.
Thus, there is a critical evidentiary distinction between the design inference made in respect to potential artifacts and the design inference made in respect to biological objects.
Therefore, the design inference in biology does not draw evidentiary support from design inferences made in respect to possible human artifacts (machines, forgeries, archaeological objects, etc.). Design inferences made in respect to artifacts may help illuminate the logical structure of such inferences, but they also illuminate key differences - the primary one being the difference between having some knowledge of a potential designer, or not.
Therefore, the design inference in biology must rest squarely in biology itself, and must be discussed in terms of biological objects and processes. Discussion of artifact situations, and analogies between biological objects and artifacts (machines, etc.), may illuminate the problems we face in making such inferences, but they don’t add evidence that biological objects are designed.
IP: Logged
|
|
Paul A. Nelson
Member
Member # 26
|
posted 17. June 2002 17:33
Charlie wrote:
if Schon is found guilty, I am quite sure it is going to be based on some sort of additional empirical evidence, not only on a statistical inference.
I agree. The investigation to look for that additional evidence is underway now. But it wouldn't even have started without the fishy matching figures.
(Incidentally, I don't even know how one would actually calculate the chances of that background noise being the same by chance - do you?)
No.
IP: Logged
|
|
Paul A. Nelson
Member
Member # 26
|
posted 18. June 2002 09:58
It is interesting to note on this web page (set up by the leading Cuban investigator of the recently discovered submarine megaliths) that "natural cause" is the first hypothesis under consideration:
http://www.medioambiente.cu/museo/exmar.htm
The intelligent cause hypothesis faces severe obstacles, not the least of which is the geological setting of these objects. Furthermore, there have been cases elsewhere of the misidentification (as intelligently caused) of naturally occurring submarine blocks.
We can expect that if the researchers claim to have found true artifacts, they will do so on the grounds of a small probability specification -- and their critics will focus on overturning or weakening that claim. We can further expect, I think, that any equivocal outcome -- i.e., "We just can't say what these objects are, or whether they were intelligently caused" -- will issue from the researchers' inability to find a small probability specification amongst the objects, and the patterns in which they are arrayed.
[ 18 June 2002, 17:11: Message edited by: Paul A. Nelson ]
IP: Logged
|
|
Paul A. Nelson
Member
Member # 26
|
posted 18. June 2002 10:17
Evan,
I disagree that there is a fundamental difference between design inferences involving a member of Homo sapiens (or some other terrestrial species [e.g., bees or beavers]) and all other design inferences. Knowing about humans as agents is extra information, helpful in our reasoning -- but having such knowledge is not necessary for inferring design.
Consider the long-awaited, yet-to-be-received extraterrestrial radio signal bearing the indicia of intelligence. If Jill Tarter and her colleagues ever do receive this signal, it will constitute the whole of their knowledge about the intelligent source. Given current technological limits on interstellar or intergalactic travel, SETI researchers may never directly observe the source(s) of any signals they receive.
But if that signal carries, say, a proof of the Pythagorean Theorem, or a representation of the Periodic Table in binary form, one would have to be a skeptic of a genuinely perverse and implacable bent to deny the reasonableness of a design inference.
I agree with you that biological design inferences need to stand on their own. But we don't need to observe, or know anything about, the postulated intelligent cause, to make such inferences.
IP: Logged
|
|
Jay
Member
Member # 268
|
posted 18. June 2002 10:32
Hi Evan,
To clear up any confusion, I wasn't implying that you were condemning IDsts or calling them said names, I was referring to the bulk of the popular science press.
I fully agree with you that analogies are only good for suspecting design, not proving it. And you are quite right in saying that other empirical methods must be used to detect it.
However, my point in writing this was more to point out the level of skepticism in biology for even *suspecting* ID from analogy. The flagellum/rotary motor is indeed a very intersting parallel to complex human design. The biological world is filled with these interacting, fine-tuned machines that look even more advanced than human engineering, if anything. So there are strong analogies (with biology often being the more advanced side of the analogy), and no good forthcoming natural forces to explain them thus far.
So, again I ask, why the obvious repulsion to the idea of even *suspecting* design in biology? Why must we remember to remind ourselves that although nature looks designed, it isn't? Why is the investigation of unguided causes to these biological machines the only game allowed in the funded scientific world? Why are scientific ID proponents like Demski and Behe treated with such scorn? I understand that we have lots of muddy sociological and historical reasons for these problems, but even still, it is clear that there is an aversion to the idea of design not found in other science areas like, say, archaeology.
And remember, in this example of the underwater shapes, we *don't* know if this was human designed or not. At this point, we're simply infering design and drawing analogies to our own designs to weigh whether we ought to suspect it.
So to close, I'm not suggesting that these striking parallels prove anything. Like you point out, analogies to design can give false positives. However, given the level of similarity to our designs, I have to wonder why even suspecting design (say, in theoretical biology journals where speculation is allowed) is taboo. What I'm getting at is that there seems to be reasons other than simple weight of evidence guiding how the scientific community treats the possibility of design. These underwater shapes seem to be a good example to illustrate just how harshly design is treated in biology vs. other fields of science.
Thanks,
jay
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 18. June 2002 20:56
Paul Nelson raised the question if we need to know more about the designer in order to infer design. Since I argue that we need to know more about the causal history of the event under study, I would argue that it surely would be helpful to know more about the designer and I am not convinced that we can infer design without making some assumptions about the designer.
For instance in the Schon case, without knowing the causal history of how the publications were prepared, it seems to me impossible to eliminate possible explanations raised by Schon himself namely that he erroneously submitted figures or that a computer glitch caused the wrong figures to be included or that a previous template was used which contained the figures in question. Without knowing how the publication was prepared it would be hard to conclude 'design' or in this case 'fraud'. Of course would not the alternatives in this case be accidental submission and would that not be categorized as design as well?
How can we reliably infer design without knowing more about the causal history and I suggest that the causal history requires knowledge about the possible designers. After all if one were to submit the various hypotheses to be eliminated, does one not need to understand the likelyhood of certain events and does that not require understanding the causal pathway?
Since the ID inference is based on elimination of alternatives does it not need to know the causal pathway? After all the probabilities of the hypotheses depend significantly on the pathway taken.
Certainly the assumption that Schon is absent minded or computer illiterate would lead to a different hypothesis outcome.
So can we continue to claim that we can infer design without any false positives?
IP: Logged
|
|
Jules
Member
Member # 181
|
posted 20. June 2002 19:13
I've lost his book, but vaguely remember a relevant part of Del Ratzsch's (sp?) book on -- and here I embarrassingly admit not remembering the exact title -- "Science, Intelligent Design, and the Supernatural." The relevant part was early in the book on what he considered "primary" evidence of design. Naturally occurring regular, geometrical shapes occur often on the microscopic and super-macroscopic levels. But on the human level -- the one most easily observed by us -- regular geometric shapes only occur by design. Finding regular geometrically shaped rocks -- monoliths -- would be strong evidence of design, regardless of who designed it -- human, alien, or supernatural. At least, so Ratzsch would maintain.
[ 20 June 2002, 19:15: Message edited by: Jules ]
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 20. June 2002 19:36
Jules:
quote:
Finding regular geometrically shaped rocks -- monoliths -- would be strong evidence of design, regardless of who designed it -- human, alien, or supernatural. At least, so Ratzsch would maintain.
Or perhaps natural?
Thanks to Paul for pointing to this link
quote:
El "Altar de la Virgen" localizado cerca de Achotal en la provincia de Guantánamo, es una curiosa estructura natural originada por el deslizamiento de grandes lajas de roca sobre el cauce de un arroyo.
quick&dirty translation:
The altar of the virgin, located close to Archoal in the province of Gauntanam is a strange natural structure that formed through the sliding of large 'lajas' of rock on the channel of a stream.
Another question is, would round rocks be evidence of design or merely monolith shaped ones? If I remember correctly granite can form interesting geometrical sapes as well.
How do we establish if these are natural or designed? I would argue that we need to know their causal history in order to assign appropriate probabilities. Is the spherical shape of the planets or stars evidence of design?
[ 20 June 2002, 19:48: Message edited by: Frances ]
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 20. June 2002 22:34
quote:
Jules:
Naturally occurring regular, geometrical shapes occur often on the microscopic and super-macroscopic levels. But on the human level -- the one most easily observed by us -- regular geometric shapes only occur by design. Finding regular geometrically shaped rocks -- monoliths -- would be strong evidence of design, regardless of who designed it -- human, alien, or supernatural. At least, so Ratzsch would maintain.
Guess he never was in Ireland. Unless of course, Giants counted as potential designers.
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 21. June 2002 07:53
I think everyone agrees that the design inference must be made against a background theory of what sorts of natural, or better, non-intelligent, processes may possibly be in play in a given situation.
This just means that the less the object in question was manipulated by intelligence, the more closely it will resemble something that might have happened naturally (without intelligence), and so the more difficult the discrimination will be.
That is not to say that clear-cut cases are not also possible. So far, we have been discussing shape alone. But what about material? I forget what the monolith on the Moon in "2001: A Space Odyssey" was supposed to be made of, but presumably not Moon rocks. So that case was pretty clear cut. (Although, even there, non-intelligent explanations might be thought up---but they would be pretty ad hoc.)
My only point is that it is seldom going to be black-and-white. It is going to require a lot of background information and judgment, and will have to be determined on a case-by-case basis.
As for life, I think everyone agrees by now that "chance" is out of the question (any dissenters?). So the question really boils down to whether a natural, law-like process was at work. And the answer at this point is, We just don't know. But that does not mean we should stop looking, IMO.
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 21. June 2002 10:52
quote:
So the question really boils down to whether a natural, law-like process was at work. And the answer at this point is, We just don't know.
That's a royal "we", isn't it?
quote:
But that does not mean we should stop looking, IMO.
If you mean "keep looking for a natural mechanisms underlying CSI-like natural phenomena for which no natural explanation is known", certainly I agree, but I thought the whole point of CSI calculations was to rule out naturalistic explanations altogether, wasn't it? If now you say CSI just rules out chance (which I don't even think it can, except in very simple cases, as the Schon example shows), we're back at square one: chance is of course not a good explanation for many natural phenomena. I have a feeling Dembski would not agree with this last statement of yours.
IP: Logged
|
|
Evan
Member
Member # 164
|
posted 21. June 2002 13:22
James writes, "As for life, I think everyone agrees by now that "chance" is out of the question (any dissenters?). So the question really boils down to whether a natural, law-like process was at work."
All histories, be they personal or evolutionary, small or large, involve a combination of law and chance. If James’ question means that everyone agrees that the “junkyard in a tornado” scenario did not happen - that one day by pure chance the components of the first cell, or flagellum, or whatever, just all coincidentally showed up in the same place and created the object in question - then yes indeed I think we all agree.
But the alternative to that is not the proposition that some inevitable working out of natural law produced life as we know it, with no chance involved at all.
I would hope that everyone agrees that law-and-chance work together in the natural world.
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 21. June 2002 19:16
Charlie D.:
Yes, you are correct. Dembski and I respectfully agree to disagree about irreducible complexity. The fundamental basis of the disagreement is the fact that I reject the assumption that organisms are machines. If we grant that assumption, then I would agree with Dembski's conclusion that a naturalistic explanation of life is impossible. As it is, I believe that Dembski's work provides a reductio ad absurdum of the premise that organisms are machines.
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 22. June 2002 14:19
In viewing the discussion on this topic I can’ help but think that there are three separate issues involved, and the opinions being expressed are often confusing the three issues. IMO, the two issues are- 1)what do the analytical results tell us about what we know and don’t know, 2)what does ‘speculation based on the analysis tell us about what we might know in the future and finally 3)why aren’t the technical conclusions of this analysis accepted by wider scientific community.
IMO, specified complexity or the sweep of chance hypothesis demonstrates that 1)observed complexity can not be explained by known processes, and 2)observed complexity can not be explained by known processes and chance. More generally, the hypothesis says that known Darwinian and neo-Darwinian processes plus chance can not explain most observed complexity. The sweep of chance hypothesis shows that there does not currently exist a scientifically verifiable explanation for many/most observed genetic/evolutionary changes. There is little or no reason to doubt the validity of this conclusion. The lack of acceptance of the analysis is not based on the analysis, but on how it is interpreted. (I’ll return to this issue latter.)
Once you have recognized what the sweep of chance hypothesis says about what we know and don’t know, then we can consider the different speculative options on what we might some day know. These speculative options include-
-unknown mechanical processes
-unknown Darwinian processes plus chance
-change produced by external intelligent forces
-unknown non-mechanical properties
-we will never now
As Dembski argues in the last paragraph of "Does Evolution have a mechanism"- QUOTE: "Science is supposed to pursue the full range of possible explanations." All the ‘we don’t know but we can speculate options should be treated equally. Once it is recognized that there is nothing in the ‘we know and can scientifically verify common’, then we can start addressing what we might some day know, or what we might someday want to know, or even ‘how to go about acquiring what we may some day want to know".
The third issue associated with the ‘sweep of chance hypothesis’ is not the facts or the technical issues but the standards used to interpret the conclusions. On a rigorous scientific basis, ‘can’t be explained by any known scientifically verifiable process or mechanism’ means ‘there is no existing valid theory’ or ‘all existing theories have been falsified’. But again to quote Dembski- "Evolutionary biology allows only one line of criticism, namely, to show that a complex
specified biological structure could not have evolved via any material mechanism." In other words, you can’t eliminate the current theory as a valid explanation unless all possible material mechanism have been eliminated as possibilities. This is clearly an irrational and scientifically unsound position. As should be obvious, Dembski’s analysis will not and can not be accepted by the profession of evolutionary biology until this unsound corruption of professional standards has been recognized and corrected.
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 22. June 2002 17:33
Warren raises some very good issues about ID and chance hypotheses.
The argument that specified complexity cannot be explained by known processes does not seem to reflect Dembski's argument. Dembski argues that it is intelligence which can generate CSI. The issue whether or not natural processes and chance can generate CSI seems largely unanswered. Dembski attempts to show this using the NFL theorem but as others have argued the NFL theorem is irrelevant and more importantly the NFL theorem likely does not apply
Warren does point out a very important limitation of the design inference quote:
The sweep of chance hypothesis shows that there does not currently exist a scientifically verifiable explanation for many/most observed genetic/evolutionary changes. There is little or no reason to doubt the validity of this conclusion.
First of all the "known processes". If that argument were logically extended then fairy circles, the movement of the planets all once were reliable indicators of design UNTIL of course science found out a scientifically verifiable explanation. So it seems clear that under these assumptions the ID inference is not the reliable indicator of design as suggested in for instancce NFL (no false positives). Since ID inherently relies on the elimination of hypothesis of the processes, it will have to deal with 'unknown processes'. Now my question to all this is: What does ID have to offer for a process? Do they provide us with a possible scientific explanation? Is such an explanation reliable to the extent proposed by ID? Of course science pursues the fullest range of possibilities and implicitly includes design by the attempts to verify /falsify known hypotheses and propose alternative hypotheses. Since ID however is based on elimination, it will fail to provide us with such a hypothesis and will have to wait until it can be shown that all chance and regularity hypotheses have been exhausted.
As far as unsound corruption of professional standards, I find this accusation unnecessarily ad hom further more Warren's conclusion that "one cannot eliminate the current theory as a valid explanation until all possibe mechanisms have been eliminated" confuses the ID hypothesis with the scientific hypothesis. All that is required is to find a better hypothesis than the current one and it will be replaced. If ID can provide us with a better hypothesis they would be able to replace the current one, but ID, being based on elimination, does not seem to have that option. It is interesting that Dembski claims that "evolutionary biologu allows only one line of criticism, namely, to show that a complex specified biological structure could not have evolved via any material mechanism".
First of all this is not what biology allows, in fact it allows for any hypothesis which is better than the current one. But Dembski's argument seems to be exactly what the ID inference claims needs to be done to infer intelligent design. So how can Dembski object to this?
As far as the claim that there is no reason to doubt the hypothesis that Neo-darwinian processes cannot explain complexity, I would like to point out that there are several papers that seem to show exactly this.
I'd propose that in order to circumvent these problems we might be starting to consider if we can indentify positive hypotheses of design?
[ 22 June 2002, 18:28: Message edited by: Frances ]
IP: Logged
|
|
|
Printer-friendly view of this topic