|
Author
|
Topic: No False Positives and the Lust for Certainty
|
Mark Szlazak
Member
Member # 391
|
posted 17. December 2002 03:23
Wesley, your right I should have used "evasive" instead of "deceptive", and you do present evidence of no attempts (bad or good) being made about the existence of a designer so far. Also, some waffling on the part of Dr. Dembski.
If this is the case then I hope things start to tighten up. Thanks.
IP: Logged
|
|
Paul A. Nelson
Member
Member # 26
|
posted 17. December 2002 11:15
John and Wesley,
Some comments:
1. According to your analysis, rarefied design -- i.e., postulating an unobserved designer to explain events -- is, in principle, inadmissible as a scientific inference.
2. Thus, the state of the evidence (the effect) is, strictly speaking, irrelevant. Unless we have independent knowledge of the cause -- i.e., the designer -- we may not postulate intelligent design as the cause of any event.
If (1) and (2) accurately represent your position, I'd like to move the discussion on to other issues.
Some side remarks:
3. Wesley, I recall that I brought up the Kreb's cycle at the CSICOP discussion, not Ken Miller. If you have the video or audio tape, could you check for me please? Thanks.
4. John, I've heard from friends in the history of science that the Laplace & Napoleon anecdote is apocryphal. What's your source?
IP: Logged
|
|
Janitor@MIT
Member
Member # 125
|
posted 17. December 2002 11:35
Actually, I thought I was using the term “rarefied” consistently with your usage and apparent intent (LOL I actually thought the term was selected to marginalize design inferences. Please correct me if I’m wrong.), and also with the signification it has for most people: rarefied = rare.
That Darwinian evolution is “rarefied” is a fair characterization, as biologists often say that with comparative rarity do they observe beneficial mutations to occur, and these are invariably simple modifications to existing systems.
Otherwise, per your stipulations, “rarefied design” plainly does not apply to life forms, except in the sense of the “unknown,” as you indicate. But I don’t find this “unknown” particularly troubling, intriguing yes, troubling no, and especially as most people are incapable of specifically identifying the designer of objects they “know for certain” had a designer. Designers typically labor in anonymity, and at most people know or infer only that they must indeed be (generically) intelligent (designers). I can personally vouch for that and more.
Basically, “rarefied design” adds nothing to our understanding here.
But like the term “rarefied design” the introduction of and identification of “intentionality” tends to distraction. I think I see a pattern developing here, and accordingly I’ve advised Messrs. Dembski, Behe, Nelson, et al, not to discuss/debate in the terms of discourse their “ideological” opponents dictate. (Not that they need or appreciate my advice. Its offered only for what its worth. LOL)
That intelligent agents or designers have intentions seems obvious enough. That there is any way scientifically to reliably abstract, analyze, and reach any firm conclusions as to what those intentions are specifically is doubtful. Even when there is no question as to the existence of intent. Obviously, designs can be modified and adapted to uses or functions for which they were never “intended.” E.g., my own mother is possessed of the rather disturbingly Luddite theory that a hammer solves all technical problems. I dare say, I don’t believe the designer, whomever or whatever (!), of the hammer intended it as a generalized problem-solver. But how would I know?—my mother may be correct! (Actually, she insists that she's always correct, and you jolly well better believe it!)
In engineering the “intelligence” of designs is studied w/o explicit reference to “intentionality” at all (per the research program begun in the early 1970’s by Saridis and independently by Albus, and even earlier by others in the field of AI and machine learning). Amazingly enough, its as if the engineers implicitly accept the existence of intent wrt design, but treat designs scientifically as if they had none. (Just as biologists treat designs as if they were not designs.) A quick review of the literature will reveal that engineers very rarely refer explicitly to their own intents, purposes, and motivations!
This may be because designers expect that “intent” is revealed in a natural and straightforward way by function. They seem to strongly, if even only implicitly, identify intent with function. There is however a problem, well-noted in the biological literature, with even identifying a “function.” This is not the case in engineering design, as “function” has an unambiguous and mathematically rigorous definition. Biologists might do well to adopt the design definition of "function." It would be useful if biologists were serious about their “intent” of reverse-engineering biological designs. They would begin to attain in their analysis the level of rigor and precision that engineers bring to forward and reverse engineering. But for biologists it is (an unnecessarily?) complicated subject.
Design predictions?—“The enormous variety of [genomic] mechanisms and circuitry raises questions about the basis for this diversity. Are these variations in design the result of historical accident or have they been selected for specific functional reasons? Are there design principles that can be discovered? By design principle we mean a rule that characterizes some biological feature exhibited by a class of systems such that the discovery of the rule allows one not only to understand known instances but also to predict new instances within the class. For many years, most molecular biologists assumed accident played the dominant role, and the search for rules received little attention. More recently, simple rules have been identified for a few variations in design. Accident and rule both have a role in evolution [just as in design engineering] and their interplay has become clearer in these well-studied cases. This area of investigation is in its infancy and many such questions remain unanswered. This review article addresses the search for design principles among elementary gene circuits (Savageau, M.A., “Design principles for elementary gene circuits: Elements, methods, and examples,” Chaos 11 (1): 142-159, March 2001).” (My interpolations are bracketed as always.)
Now any kind of design logic, rule, protocol, prescription, or intent would seem to be excluded a priori on Darwinian grounds. Instead, all we are left with is it “just happened.” That’s funny. (My apologies, Mr. Moderator. I just can’t seem to help myself when it comes to parting cheap shots. As usual I’ll leave it to your discretion.)
[Amusing footnote:
http://arxiv.org/PS_cache/adap-org/pdf/9910/9910002.pdf]
[ 17. December 2002, 14:02: Message edited by: Janitor@MIT ]
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 17. December 2002 13:34
John: Each is "designed" for a suite of tasks, and these tasks conflict. There is no way you can extrapolate from the adaptations of a gazelle to escape lions and cheetahs, and the adaptations of lions and cheetahs to bring down gazelles, to a single designer for both sides. On purely empirical grounds, you must infer a separate designer for each organism (which rather undercuts the rationale for the designer hypothesis in the first place, I think).
Really? Let's assume that the human body is designed. Osteoblasts and osteoclasts are "designed" for different, conflicting tasks. Please explain why it is that someone who thinks the human body is designed must infer at least two designers.
IP: Logged
|
|
Micah Sparacio
Member
Member # 6
|
posted 17. December 2002 15:13
Along with Mike Gene's remarks... it is important to note that engineers regularly face optimization hurdles in which the optimization of one feature leads to the degradation of another feature. Theoretically, optimal design will allow for the ultimate optimization of each feature in the system without negative side effects. However, as human history shows us, design regularly involves the *balancing* of multiple conflicting features in such a way that the overall system maximizes its function.
Why then should we expect that competing design features undermine the design hypotheses?
IP: Logged
|
|
Dene Bebbington
Member
Member # 168
|
posted 17. December 2002 17:09
Micah wrote:
>However, as human history shows us, design
>regularly involves the *balancing* of multiple
>conflicting features in such a way that the
>overall system maximizes its function.
Are humans designed? If you believe so then what do you consider their function to be?
--
Dene
IP: Logged
|
|
John Wilkins
Member
Member # 418
|
posted 17. December 2002 22:15
Janitor: Here is the on-line Merriam Webster definition:
Main Entry: rar·e·fied
Variant(s): also rar·i·fied /'rar-&-"fId, 'rer-/
Function: adjective
Date: 1941
1 : of, relating to, or interesting to a select group : ESOTERIC
2 : very high
I was not aware of the first definition, but it is somehow apposite. Here are some other online definitions. I had in mind the last.
Paul: I will check my sources (I tutored in History of Astronomy the past two years, but my references were a bit old). However, even if apocryphal, the point remains.
[ 19. December 2002, 01:31: Message edited by: John Wilkins ]
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 17. December 2002 22:38
quote:
However, as human history shows us, design regularly involves the *balancing* of multiple conflicting features in such a way that the overall system maximizes its function.
Why then should we expect that competing design features undermine the design hypotheses?
So, what is the function of the gazelles vs. lions arms race? Or, immune systems vs. diseases?
Here's MHO: if one is going to take the "obvious" design in biology at face value, i.e., "someone designed flagella for swimming", then one must also take the similarly "obvious" battling designs and arms races in biology at face value, as being either due to a single designer subverting his own designs (through e.g. insanity, perversity, or ineptness), or to multiple designers battling each others' designs. For some reason these kinds of possibilities are never taken seriously as hypotheses by the ID crowd.
One can avoid the "obvious" conclusions by having (a) a more complex hypothesis (lions and gazelles serve function X, except in Madagascar and numerous other places where they could live but don't) or by (b) giving natural selection the credit for such instances of creativity, but then the design hypothesis is either (a) no longer a simple, obvious inference or (b) completely superfluous.
IP: Logged
|
|
John Wilkins
Member
Member # 418
|
posted 17. December 2002 22:48
What yersinia said. Why can't I be so concise?
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 18. December 2002 00:55
Yersinia,
I for one have never considered any of my design inferences to be simple or obvious. After all, recall that the "obvious" thing for a designer to do was supposedly to strip cytosine from DNA. Yet apparently, the designer was smarter than this. I don't mean to sound offensive, but non-teleologists don't make very good design theorists.
It would seem to me that your argument imposes a narrow creationist perspective on ID, where the designer, for example, directly designs gazelles and lions de novo. Yet there are other options for the telos. For example, what do gazelles vs. lions and parasites vs. hosts have in common?
You write: For some reason these kinds of possibilities are never taken seriously as hypotheses by the ID crowd.
I've explained my reasons for not taking these very seriously - they appear to be fruitless ideas and are not well thought out. As just one example, consider a point I raised back in October:
quote:
I still don't see the evidence that is more consistent with multiple designers than with a single designer. The evidence RBH cited was as follows: contrary designs; complementary designs; imperfect designs; designs over time. I don't see why any of this would be more consistent with MDT than SDT. In fact, if all of this is supposedly more consistent with MDT, what is left that would be more consistent with SDT? Apparently, a single, perfect design event???
If one wants to redirect the focus to a designer-centric approach, and thus seriously raise multiple designers, you need to explain what you would expect from a single designer. Thus far, you seem to think that design behind life would not include any predation and any parasitism. Is that all?
[ 18. December 2002, 00:56: Message edited by: Mike Gene ]
IP: Logged
|
|
Wesley R. Elsberry
Member
Member # 122
|
posted 18. December 2002 01:25
Paul,
quote:
3. Wesley, I recall that I brought up the Kreb's cycle at the CSICOP discussion, not Ken Miller. If you have the video or audio tape, could you check for me please? Thanks.
My VCR has taken up the habit of eating tapes, so I won't be putting in my copy until I get a new VCR.
I couldn't say from recall that you didn't mention the Krebs cycle first, since you preceded Ken, but I do recall that Ken specifically discussed the Krebs cycle as a counterexample to Dembski's EF/DI. That's because Ken used my "origination probability calculator" to "do the calculation", putting his finding on a par with the discussion by Dembski of the E. coli flagellum. The same equations were used in both cases. (Ken's presentation even included a slide with a screenshot of my web page. More of Dembski's equations are implemented on my Finite Improbability Calculator.)
I recall that you disputed Ken's use of the Krebs cycle on the grounds that the evolutionary scenario Ken cited was not complete enough to satisfy you. That was during the panel discussion. If I find my audio tapes, I'll give them a listen.
Wesley
Archived at this page.
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 18. December 2002 01:38
Does this mean that teleologists don't make very good natural mechanism theorists?
Let's rememember that your assertion, while interesting seems to be somewhat ad hoc. Your use of the term 'designer' seems to include natural processes btw. I so far have found that applications of 'design' seem to be limited to 'regularity or chance' cannot explain it or it resembles design used in 'engineering' (feedback loops, robustness etc).
Yersina also seems to agree with your statement about 'obviousness' in fact, so far design may be only obvious in the eye of the teleologist. In order for design especially when addressing rarefied design to show any relevance beyond helping us to understand how something works (analogy, metaphor) one would need to explain what design really predicts in a non ad hoc manner. What does teleology contribute to our understanding? Why is it needed? Front loaded design, which is what you seem to be alluding to, seems so far indistinguishable from natural processes, so why the need for a design inference?
I do agree that if the SDT wants to gain any relevance it does need to explain what one would expect from a single designer. Of course one can always move the goals posts to match whatever data we have but what does a single designer then contribute? Why not use the metaphor of multiple designers? Seems to match reality much better and perhaps would explain the data much better?
So in other words why should ID reject in advance multiple designers?
I could think of some reasons but none would fall in the category of science.
So far serious efforts to look at 'single' designers seems to suffer from the same problem you have identified for 'multiple designers', a lack of a non ad hoc model.
[ 18. December 2002, 07:38: Message edited by: Moderator ]
IP: Logged
|
|
John Wilkins
Member
Member # 418
|
posted 19. December 2002 00:35
Response to Paul:
First, the issue about Laplace: I found the discussion in Toulmin and Goodlfield's Fabric of the Heavens but without a citation. I have read it in more detail elsewhere but cannot find it. A Rupert Hall's A Brief History of Science discusses the more general aspects of Laplace's views.
Now, you asked
quote:
1. According to your analysis, rarefied design -- i.e., postulating an unobserved designer to explain events -- is, in principle, inadmissible as a scientific inference.
2. Thus, the state of the evidence (the effect) is, strictly speaking, irrelevant. Unless we have independent knowledge of the cause -- i.e., the designer -- we may not postulate intelligent design as the cause of any event.
If (1) and (2) accurately represent your position, I'd like to move the discussion on to other issues.
I should like to make a final clarification - I do not think an unobserved designer is an illicit inference. An unobserved and unobservable designer - that is, one about whom we have no independent information and are not likely to - is. I can infer that this garden was designed if the output is in keeping with what I know of both ungardened plots and other gardening. But if I have no way to tell the difference between an ungardened plot and a subtly gardened plot, and no information about the designers who might build such (say, for a movie set), then I cannot reasonably infer a designer. But if I find that the plot has no "back" to it (i.e., it's a painted set, or the "plants" are made of linen) then I still might be able to infer quite a lot on the basis of prior information about film makers, etc.
But suppose that there is no observable difference other than what someone speculates might be in place in the absence of a gardener - suppose that I tell you, with no supporting evidence, that this stretch of bushland would look like a dusty plain without a gardener - why would I be forced to conclude that there is, in fact, a gardener at all?
[Those who have read Anthony Flew's retelling of John Wisdom's parable in "Theology and falsification" in New Essays in Philosophical Theology will recognise some of this.]
So, if I have information about gardeners and how they behave and what they make (and even, perhaps, why, although that is not necessary), I can make an (ordinary) gardener inference even in the absence of observing that gardener at work. But you ask for a rarified gardener inference...
[ 19. December 2002, 01:36: Message edited by: John Wilkins ]
IP: Logged
|
|
Paul A. Nelson
Member
Member # 26
|
posted 19. December 2002 09:26
Thanks for your clarifications, John. About Laplace: I looked around my office, and every source that mentions the Napoleon story (e.g., E.J. Dijksterhuis's The Mechanization of the World Picture [1961, p. 491]) puts it in terms of "it is reported that" or "it is said that" -- but never with a citation to the primary report. It would be a shame to ruin such a pithy illustration with evidence that it really happened. ![[Wink]](wink.gif) Incidentally, I take the point of the story (apocryphal or not) to be that theories of divine action, where they involve observable states of affairs, are empirically vulnerable. Presumably, Newton might have been right about the instability of the Solar System.
Wesley, John has endorsed my two summary points of your (joint) position. Can I get an amen from you? I'd like to move on to more central issues.
[ 19. December 2002, 13:47: Message edited by: Paul A. Nelson ]
IP: Logged
|
|
Wesley R. Elsberry
Member
Member # 122
|
posted 19. December 2002 20:57
I agree with John's clarification.
So, what are the central issues? I'm thinking those are the issues from Dembski's topic-opening post.
The title of the thread talks about a "lust for certainty". I think I've established that the claims of certainty can be found in Dembski's work. That critics chose to respond to those claims should have been expected. The more tentative claims came later and did not displace the non-tentative claims, leading to a state of inconsistency in Dembski's claims.
The claim by Dembski that there exist no counterexamples to his claim that his EF/DI is "reliable" is either abysmally weak (based on less than a handful of worked examples) or actually false (since various critics have put forward candidate counterexamples).
I've had a look at Demsbki's essay on "logical underpinnings" but I don't yet wish to comment. I'm getting acquainted with BibTeX at the moment.
Wesley
Archived as usual. (The other links work for this post, too.)
IP: Logged
|
|
|
Printer-friendly view of this topic