|
Author
|
Topic: The Evolution of the Bacterial Flagellum
|
Pim van Meurs
Member
Member # 541
|
posted 07. December 2003 15:16
Nelson: And so design seems to be a good explanation for it, even if we have no idea exactly how it happened. It's just something you're accepting for the moment to see where it leads.
Design may indeed be a valid observation but that does not help us determine the nature of the designer(s). The whole issue of biology and design is that design in nature is not really the issue. Regularity and chance processes under selective forces will inevitably lead to structure containing information and complexity. And such processes will also appear to be teleological in nature.
So the question is not design in nature but the nature of the designer(s).
Nelson: For some scientists the DNA looks like rubbish, the work of a sloppy tinkerer and so a sloppy beginning seems to be a good explanation for it, even if they have no idea how it happened. For other scientists DNA looks like a miracle, a work of perfection.
And to most scientists DNA looks like a combination of a sloppy tinkerer leading in some cases to sub optimal solutions and in other cases to solutions closer to optimal.
Scientists look for explanations for these observations, propose pathways, rule out pathways and identify 'designers'.
Perhaps we can place these pathways, even imperfect ones, head to head with ID proposals and determine which ones fare best?
[ 07. December 2003, 15:19: Message edited by: Pim van Meurs ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 07. December 2003 15:21
Ged
But I see that you avoided my question by changing the subject.
Which question did I avoid?
Ged
But the EF amplifies any unknown distributions just as much as it amplifies the known "chance" distributions by the factor of M*N from "ProbRes" factors.
How so?
Ged
As to OOL research, that is entirely about theory of how different sequences of events could relate. Of course I have no way to compare the probability of space ships coming down to deposit first life, and relationships being researched in OOL theory. I don't know if that is high, or low probability, so I have no way to make a comparision.
Saying that something looks designed has nothing to do with the "probability of spaceships", it has to do with what properties the system has.
Even if we could somehow calculate the "probability of spaceships" that has nothing to do with "coincidence" in the case of the EF. "Coincidence" in the EF has to do with a chance occurance, such as attributing reversion to adaptive mutatgenesis, when in fact it's just because some amplifications include dinB, whose gene just happens to lie near lac.
What is being observed is a property of the system. Furthermore, the EF does not consider design hypothesis as "chance hypothesis". So there is no need to calculate any "probability of space ships". Note Sober himself admitted this:
quote:
To infer watchmaker from watch, you needn't know exactly what the watchmakers had in mind; indeed, you don't even have to know that the watch is a device for measuring time. Archaeologists sometimes unearth tools of unknown function, but still reasonably draw the inference that these things are in fact, tools.
[ 07. December 2003, 18:30: Message edited by: Nelson-Alonso ]
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 07. December 2003 15:43
Mike:
Indeed. It’s a Darwinian perspective , a way of looking at things. And yes, it does have certain epistemic advantages (as I state above). But is there more?
That is largely a philosophical question it seems.
Mike: I’m not concerned about any ID “cause,” so there is no need to be concerned for me. As far as your own opinions go here, I haven’t the foggiest idea of what you would consider “evidence relevant to ID.” Thus, I’m not sure I can help you here.
I am merely pointing out the obvious fact that an ID proposal which relies fully on regularity and chance seems to be lacking in support in favor of ID.
Mike then moves away from ID to the usefulness of ID premises.
quote:
At least we can reach agreement on the plausibility of designing through evolution. I’ve explained the logic of this position several times. Does it have scientific relevance? I have found it to be so. It provides a frame of reference that sparks inquiry (for example, the closer look at cytosine deamination in relationship to the genetic code and side chain hydrophobicity). The frame of reference is subtle and is easy to overlook if one is demanding a sensational proof of design.
First of all lets establish that there is a subtle switch here from infering ID to using teleology in defining hypotheses. As history has shown, biology is full of teleology and scientists have used teleological thinking quite productively.
So I do not doubt that Mike's teleological approach to Cytosine can be useful in helping us understand biology.
BUT there is an immense difference between usefulness of teleology and intelligent design approaches and the need for intelligent designers. Since selection in nature will give the appearance of teleology, evolutionary science has identified a powerful designer in nature which can (in principle) explain much of the observed data.
Mike: You didn’t answer my question. Were those real evolutionary pathways? Are you saying the F0F1 ATP synthetase is indeed homologous to the TTS machinery?
What I am pointing out is that evolutionary scientific approaches have been far more succesful in proposing pathways for the flagellum and that I am very interested if ID relevant pathways have been proposed that could be compared with other hypotheses?
Mike: Science already has limitations and I am pointing out that the assertions of the ID critic entail yet another limitation.
It is not the assertions of the ID critic that entail another limitation but rather the approaches sought by ID proponents. Cause and effect can be easily confused in these concepts.
Mike: Say the bacterial flagellum is indeed designed. If the critics are correct, the only way science could ever detect this truth is if it first uncovered data about the flagellum’s designers.
Just like science we need to explain the pathways. The problem with ID is that it is unconstrained and thus useless as a scientific explanation. What about pink fairies, the unicorn, Zeus and company, LGM from Mars... All valid ID 'explanations' but what do they explain? Until ID can constrain its explanations, no explanations really exist. Worst case we have a scenario of 'we don't know' (See Elsberry and Wilkins "Theft over toil" paper for instance)).
Mike: I’m simply pointing out that you have no idea what science would look like if ID were true. Aside from the fact that this severely weakens your ability to constructively critique an ID hypothesis, it is understandable given the manner in which you strongly link your (potential) ID thinking to designer-centric demands.
This is of course a total non sequitur. The validity of scientific criticism exists independently of the beliefs of those providing the critique.
It is up to Mike and other ID proponents to show us what science would look like if ID were true. So far the proposals seem to fall quite a bit short of this and it does not take knowing what science would look like if it were true, to identify these shortcomings. An approach like Mike's which fully embraces natural processes suggests that ID fully accepts regularity and chance processes. At most it can argue that its use of teleological thinking helped us understand these processes. How is this relevant to ID ? Mike may want to explain this.
It seems to me that Mike is arguing, and he can correct me if I am wrong, that ID critics cannot contructively criticize ID hypotheses. If that were the case science would since long have stopped.
Mike: I’m not sure what this had to do with my point - you seem to be arguing that yes, science is completely dependent on knowledge about the designers in order to infer design, but no, this is not a limitation of science. If it is a fact that a design inference depends on independent knowledge of the designers, then this is another limitation of science.
Absent such knowledge science can infer at most 'we don't know'. What would science be without any independent evidence. It could propose totally unobserved pathways to explain past data. A fascinating form of speculation without much relevance to help us focus on what really happened.
Of course a miracle could have happened, of course intelligent design could have happened, that is not the issue. An explanation which explain anything explains nothing.
I also thank Gedanken for his thoughtful responses which mimic to a large extent my intent.
quote:
Without any limitations on that technology, there is no difference in terms of the unconstrained and unknown nature of the prior joint probabilities of the designer effecting the event--thus there is no way to pinpoint the reliability of an "explanatory filter" like inference.
and
quote:
Yes, I am waiting for the modest, systematic, and positive presentation of the means, motive, and opportuntiy for human-like intelligence to design the flagellum.
It would be helpful if these issues were addressed.
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 07. December 2003 17:56
Okay, I couldn’t resist jumping back into this fray.
Gedanken: Yes, the problem is what are the limitations on "human-like" designers. Are they limited to being present and having access to the events that occurred?
Yes. I envision the intelligent intervention behind the design of life and its seeding. This is a modest extension of the hypothesis that Crick and Orgel published. Afterwards, it’s a question of front-loading, where we can study how design exerts itself, over time, without the designer being present.
Or are they allowed to have technology far beyond anything we have ever seen in order to implement their "designs".
Here we have to tease apart espistemic concerns from ontological issues. From an epistemic perspective, fantastic technology poses the very problem you cite. But from an ontological perspective, it is absurd to demand that any designer of life would have to possess human technology capabilities circa 2003. If you insisted on this standard, we can use philosophy to rule out design. That is, since humans can not design homeostatic, self-replicating cells, then it would be impossible for any intelligent agent to do likewise. This strikes me as an irrational position. The rational thing is to recognize that other possible intelligent agents may possess a knowledge and technology base that far surpasses our current state. But that takes us back to the epistemic concerns.
However, if we do constrain ourselves to human-like designers, this is not the problem you think it is. Any technology that is so fantastic by our standards is going to lack a decent analog. So I don’t factor things like poofing or time travel into the picture. Instead, what I envision are reasonable extensions of our capabilities in programming, biotechnology, nanotechnology, and engineering. This is where one of my ARN replies comes into play. When asked what type of research I envision from the hypothesis that life was designed, I explained:
quote:
The best way to understand design is to do it. We’ve learned much about aerodynamics as a consequence of designing planes. I recall watching a documentary about some engineer who sold the military on plans for a type of flying saucer. Stuff worked great on the blackboard. But when they actually designed it, they couldn’t get it to fly.
Attempts to engineer homeostatic, self-perpetuating cellular systems would help us get a much more rigorous handle on the design principles involved. And that we would not simply borrow from LAWKI (life as we know it) would give us greater confidence we were looking at general principles defined by the design problem. Take for example Art’s assertion that “abundance = importance” is an engineering principle. With actual experience of designing such cellular systems, such a claim can be refuted or validated as it relates to design and life. Or better yet, it may apply under certain parameters and defining these parameters would be of use. If a rigorous principle is uncovered, one then turns to LAWKI to determine how well the design principle maps to it. The idea is simply to get a better handle on design as it relates to biotic phenomena (our current use of biotechnology is little more than primitive tinkering) and then to see how they integrate with life.
Even better than all this is the study of such designed cells over time. That is, how do they evolve? What would the descendents of an originally designed stem population really look like? We might even address issues of interest at the moment, such as, for example, the hypothetical existence of RNA-based cellular life forms. We do not even know if RNA-based cellular systems could survive for more than a few generations without collapsing under the load of error. Let’s put it to the test. What would be even more interesting is to put FLE to the direct test. Can you design evolution to reach certain targets over time? In a sense, the ability to design life transforms ID into science. And the less our life forms are dependent on knowledge/material from LAWKI, the more significant the similarities.
Gedanken: This is an area of possible conflation--in which Mike makes us believe that he has limited the field to "designers" that are realistic, like humnans who were physically present, while in fact the limitation is only of a very insubstantial nature of some limited areas of constraints of how they behave, with no constraints on their technology to achieve their "goals".
Once one begins to understand my hypothesis (Front-loading through Seeding), it should become clear that the whole hypothesis is built around such constraints. If there was no constraint envisioned, Rex’s “look at more recent systems” argument would resonate more.
Without any limitations on that technology, there is no difference in terms of the unconstrained and unknown nature of the prior joint probabilities of the designer effecting the event--thus there is no way to pinpoint the reliability of an "explanatory filter" like inference. If you allow for "human like" designers to swoop in at arbitrary times with their space ships, how have you limited the prior probabilities in any way based on knowledge? Any fantastic technology is allowed, just as in omnipitent designers any fantastic way of being all knowing and all causing is allowed. There are no constraints, even though the wording makes it sound like reasonable constraints are being offered.
And what have I written here, on my web page, or on ARN, that entails arbitrary swooping? It’s not a question of wording or sound. It’s a matter of record - my design approach is so constrained that the most common complaint is that it is indistinguishable from traditional evolutionary views! ![[Smile]](smile.gif)
Gedanken: But Mike, police work is based in large part in establishing motive, means, and opportunity. Miss out on any one of the three, and one has not drawn any reasonable inference.
But that’s not relevant to the argument I was addressing – the need for competition from the beginning. I was simply pointing out that you don’t need two competing suspects to carry out an investigation.
Your methods concentrate on motive, and leave means and opportunity as free variables, inelligible for study.
The means is not truly a free variable (see above). And as also explained above, as our own experience with biotech and nanotech increases, this variable will tighten. As for opportunity, yes, this poses a problem since it is completely dependent on independent knowledge of the putative designer. Yet access to this knowledge is not entailed by the truth of design.
Study them and problems occur with your methods.
Of course there are problems. One adjusts for this by adjusting the reliability of the conclusions. I propose only tentative hypotheses, not claims to have proved something.
There is no doubt in my mind that detecting design without the luxury of an independent base of knowledge about the designers is a difficult task indeed. I have found that many are itching to abandon the effort at the first sign of difficulty. Given my high tolerance for ambiguity, I am not bothered by the things that bother others. I just keep on probing, pondering, testing....like a toy bunny.
What is essential is having theory of how the event occurred with sufficient detail to eliminate or substantially reduce the effects of coincidence. Without theory of some sort to limit all of motive, means, and opportunity in police work, one would detain suspects on whim.
My own survey of my own approach is that it is too focused to be whimsically tagging suspects. And it gets better with time, thought, and practice.
And there is an issue as well on just what "investigation" means. Mike might say that he is only "investigating" the design of flagellum. But what becomes serious investigation? When President Kennedy was shot, did all Republicans become suspects? Certainly not! Did we say, "Oh, there could have been space ships delivering shooters to the grassy knowl"? No, in police work we limited the investigation to people who existed, at least in "theory".
The truth of the flagellum’s design does NOT entail we’d have access to a base of knowledge about its designers. Why would you hold an inference hostage to facts about X that don’t follow from the truth of X?
Yes, I am waiting for the modest, systematic, and positive presentation of the means, motive, and opportuntiy for human-like intelligence to design the flagellum.
Which is convenient for those who know we have no technique for catching and studying the hypothetical designers (the only way to gather valid information about such things). Of course the designer-centric approach is doomed to failure if we attempt to ponder things like the design of life. That’s the whole impetus for feeling out alternative approaches. Those who insist the designer-centric approach is the only possible valid method are free to declare its inability to detect life’s design. What I don’t understand is why people would take an approach and insist we all apply it to an area where it cannot work.
[ 07. December 2003, 18:01: Message edited by: Mike Gene ]
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 07. December 2003 19:01
Mike: Yes. I envision the intelligent intervention behind the design of life and its seeding. This is a modest extension of the hypothesis that Crick and Orgel published. Afterwards, it’s a question of front-loading, where we can study how design exerts itself, over time, without the designer being present.
But this form of "intelligent design" seems to be not much different from regularity and chance. We already know how evolutionary processes could give the appearance of front loading. For instance lets assume that Caporale's hypothesis is correct that selection is responsible for a bias in mutations, causing beneficial mutations to occur more often. It is tempting to see such 'front loading' as evidence of intelligent design but in this case front loading may be well understood in terms of regularities and chance and the designer(s) may be identified.
So unless we can envision ways to distinguish front loading through intelligently designed seeding versus non intelligently designed seeding, an ID based hypothesis value may be limited to being helpful in the sense that teleological thinking may be helpful to understand the natural underlying processes.
Mike: Of course the designer-centric approach is doomed to failure if we attempt to ponder things like the design of life. That’s the whole impetus for feeling out alternative approaches. Those who insist the designer-centric approach is the only possible valid method are free to declare its inability to detect life’s design. What I don’t understand is why people would take an approach and insist we all apply it to an area where it cannot work.
It seems to me that Mike is arguing that for instance eliminative approaches to ID may have to be dropped in certain circumstances? This seems to mimick Del Ratzsch observations that the eliminative approach may be quite unsuitable in detecting new design?
But I would like to extend Mike's argument to include design centric approaches which fail to propose hypotheses which can compete with non-ID design proposals? In fact when such approaches seem to embrace chance and regularities fully and leave the concept of design open as an initial condition, we still have no scientific way to resolve the issues of who/what designed? Is a design centric approach which uses front loading not subject to the same failures as a designer centric approach for such events?
What does an initial condition tell us about how this initial condition was reached? And combined with an evolutionary theory of chance and regularity, how can we establish if the initial condition was front loaded in the sense that it was (final) goal oriented or that evolutionary mechanisms used the available variation and mechanisms and tinkered succesfully to achieve a much limited teleological function?
What would be helpful is a theoretical foundation which would help us resolve these real issues. Unless we have access to mechanisms, pathways, motives etc, I doubt that science can reach more than a conclusion of 'we don't know' for such rarefied design instances. Perhaps Wilkins and Elsberry's paper can be used as a foundation for the development of such a theoretical approach?
[ 07. December 2003, 19:03: Message edited by: Pim van Meurs ]
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 07. December 2003 21:45
Pim: I am merely pointing out the obvious fact that an ID proposal which relies fully on regularity and chance seems to be lacking in support in favor of ID.
Why is that? You seem to be arguing that Intelligent Design can only express itself by violating the laws of nature.
First of all lets establish that there is a subtle switch here from infering ID to using teleology in defining hypotheses.
Sure. Yet the tentative inference to design can be a powerful impetus for pushing teleology to its limits.
As history has shown, biology is full of teleology and scientists have used teleological thinking quite productively.
This is yet another example where the critics of ID don’t all agree, as some argue that teleological thinking has been quite unproductive. I agree with you on this one.
BUT there is an immense difference between usefulness of teleology and intelligent design approaches and the need for intelligent designers.
Why make “the need for intelligent designers” the fulcrum? If the “need for intelligent designers” becomes central, then the demand for a proof of the impossibility of evolution is right behind. Perhaps it’s not a question of needing intelligent designers, instead…..
Since selection in nature will give the appearance of teleology, evolutionary science has identified a powerful designer in nature which can (in principle) explain much of the observed data.
…perhaps there are instances where intelligent designers better account for this appearance of teleology. We shall see.
What I am pointing out is that evolutionary scientific approaches have been far more succesful in proposing pathways for the flagellum and that I am very interested if ID relevant pathways have been proposed that could be compared with other hypotheses?
What’s an “ID relevant pathway?” That sounds like FLE to me, but you dismiss this approach because it does not invoke miracles and things regularities and chance could not possibly explain. If I turn to intelligent intervention, there are no “pathways,” there are recipes, protocols, blueprints, etc. What then interests you has been carefully chosen. ![[Wink]](wink.gif)
It is not the assertions of the ID critic that entail another limitation but rather the approaches sought by ID proponents. Cause and effect can be easily confused in these concepts.
The assertions are as follows: science requires an independent base of knowledge about the designers to infer design. This is then a limitation of science.
Just like science we need to explain the pathways. The problem with ID is that it is unconstrained and thus useless as a scientific explanation. What about pink fairies, the unicorn, Zeus and company, LGM from Mars... All valid ID 'explanations' but what do they explain? Until ID can constrain its explanations, no explanations really exist.
While there is an obvious connection between intelligent engineers and things like machines, I don’t see the connection between pink fairies, unicorns, Zeus, or LGM and machines. Unless of course, you want to envision such entities as intelligent engineers, in which case, their pinkishness, fairyishness, unicornishness, Zeusishness, littleishness, greenishness, and mannishness are all irrelevant.
This is of course a total non sequitur. The validity of scientific criticism exists independently of the beliefs of those providing the critique.
Not usually. If one is conditioned to investigate in a non-teleological and designer-centric manner, expecting ID to present itself as a sensational finding that is completely inexplicable by such approaches, chances are that the criticisms will be off target. I have commonly found this to be the case.
[ 07. December 2003, 21:46: Message edited by: Mike Gene ]
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 07. December 2003 23:38
Pim: I am merely pointing out the obvious fact that an ID proposal which relies fully on regularity and chance seems to be lacking in support in favor of ID.
Mike: Why is that? You seem to be arguing that Intelligent Design can only express itself by violating the laws of nature.
Are you saying that Dembski's approach to intelligent design requires a violation of the laws of nature? Fascinating.
But I am not arguing for a violation of law and nature, I am arguing that an explanation which can be fully captured by regularity and chance, or evolutionary mechanisms, such as you seem to be arguing with your front loading examples, seems to not present much evidence in favor of intelligent design. On the other hand when the intelligent designer becomes a rarefied concept we can use it as a placeholder for "we don't know" although I personally would prefer admitting our ignorance.
Mike: …perhaps there are instances where intelligent designers better account for this appearance of teleology. We shall see.
Certainly front loading does not seem to be able to resolve this? Without at least something more than appeal to fully natural processes after the 'initial condition'?
What I am pointing out is that evolutionary scientific approaches have been far more succesful in proposing pathways for the flagellum and that I am very interested if ID relevant pathways have been proposed that could be compared with other hypotheses?
Mike: What’s an “ID relevant pathway?” That sounds like FLE to me, but you dismiss this approach because it does not invoke miracles and things regularities and chance could not possibly explain.
If Mike wants to argue his own strawman despite the fact that I have made it more than once clear what I mean by my statement, that's fine with me. I guess dealing with my question about ID relevant pathways was too complex to be handled in a more constructive manner?
FLE does not sound like an ID relevant pathway to me other than the 'begging the question' assumption front loading requires intelligent design. Either the evidence shows that 1) at instance before the front loading we still see evolutionary pathways leading up to the instant of suggested front loading 2) data remains lacking about what happened before this initial condition 3) we find evidence that the front loading was in fact introduced by intelligent designer(s). So do we have evidence of 3) ? I already have argued that evidence of 1) can exists. Arguing that the evidence is compatible with an intelligent designer is meaningless if we cannot constrain the intelligent designer. If the argument is that the evidence is compatible with intelligent design then per 1) natural processes seem to not be able to be eliminated from intelligent design inferences.
Pim: It is not the assertions of the ID critic that entail another limitation but rather the approaches sought by ID proponents. Cause and effect can be easily confused in these concepts.
Mike: The assertions are as follows: science requires an independent base of knowledge about the designers to infer design. This is then a limitation of science.
Science needs something that limits the explanatory power of invoking an intelligent designer. That science has to admit ignorance sometimes when the evidence is lacking is not a problem of science but a problem with our ability to collect the evidence. If ID has to continue to rely on such approaches then ID may have chosen the wrong paths.
Just like science ID needs to explain the pathways. The problem with ID is that it is unconstrained and thus useless as a scientific explanation in at least its present format.
Pim: What about pink fairies, the unicorn, Zeus and company, LGM from Mars... All valid ID 'explanations' but what do they explain? Until ID can constrain its explanations, no explanations really exist.
Mike: While there is an obvious connection between intelligent engineers and things like machines, I don’t see the connection between pink fairies, unicorns, Zeus, or LGM and machines.
A bit of a jump in logic here. What machines are we talking about. Certainly the use of analogies between biology and machinery is only superficially relevant. Which is exactly why fairies seems to be as well an explanation for intelligent design that is unconstrained or rarefied as any form of intelligent design.
In the end what we may have to conclude is that rarefied design cannot and should not be used as evidence of intelligent 'designers'.
Pim: This is of course a total non sequitur. The validity of scientific criticism exists independently of the beliefs of those providing the critique.
Mike: Not usually. If one is conditioned to investigate in a non-teleological and designer-centric manner, expecting ID to present itself as a sensational finding that is completely inexplicable by such approaches, chances are that the criticisms will be off target. I have commonly found this to be the case.
Perhaps what you have found is that those familiar with scientific inquiry can quickly identify the major shortcomings and limitations in your for of 'ID hypothesis'? In fact much of the criticism of your ideas that I have seen seems to be quite on target. If Mike wants to continue to create a strawman argument for how critics approach Mike's ideas then little progress can be made (I am refering to his continuous reference to the supernatural). If however Mike can rise to the occasion and address the very real critiques of the shortcommings of his approaches then we may be able to see if there is something truely relevant to ID to be found. If ID has to resort to, it's just like evolution says but there may a hidden designer(s) be present that explain the initial conditions, then ID seems to have little to add to scientific knowledge beyond showing that teleological thinking can sometimes serve its purposes.
Note that Mike also shifted the argument from Mike suggesting that a critic's inability to envision what an ID world would look like would make such a critic less able to construct a valid criticism of ID to "I have seen how some people are unable to accurately address the issues raised by Mike since they, according to Mike at least, require sensational findings". It's hard to track these moving goalposts. It all begins with accurately describing the critic's arguments imho.
Although it may sometimes be tempting to dismiss criticism with a single swoop argument of 'they don't know what ID is all about', making ID even less a scientific endeavor as its success or failure seems to depend on one's subjective views.
If Mike wants to address for instance my comments on front loading and Caporale which show clearly how the idea of front loading is merely replacing our ignorance to a different place. If ID cannot deal with rarefied design then it should conclude what science would conclude namely "We don't know (yet)".
So Mike, rather than suggesting that I am proposing a supernatural requirement for intelligent design when I use the commonly used ID terminology of regularity and chance, lets focus on the real issues raised here. If your variant of ID is nothing more than an initial condition followed by fully natural processes not requiring any further interventions then you merely argue that evolutionary science is correct at least from the moment of front loading forward. The actual event of interest seems to remain shrouded in our ignorance at most.
In fact I would argue that most if not all evolutionary sequences will give the impression of 'front loading' since evolution tends to (re)-use existing genetic information.
The vaste evidence of common descent supports this. Any 'front loading' which failed to deliver will have been mostly eliminated by selection and time thus remaining are evolutionary succesful pathways. That such pathways tend to appear front loaded complicates matters when for instance taking into consideration that front loading appears to have happened countless times throughout history, unless Mike can propose which front loading is ID relevant and which front loading is not? For instance front loading introduced by biased mutations ala Caporale is argued to have evolved. If Caporale is correct then we have another example in which 'front loading' is insufficient to resolve the issue of intelligent design and intelligent design merely serves as a place holder for 'we don't know' or 'initial condition' neither one seems to be particularly relevant to the common usage of the term intelligent design.
Having said this, I do agree that a fruitful co-existence of science and religion depends on front loading ideas. That God set in motion a complicated sequence of events with the Big Bang is both comforting to me as a Christian and acceptable to me as a scientist.
Some examples from Mike's own website
[/b]
(1.) First, it would appear more and more that a designer could design mutlicellular life through unicellular life forms. This is front-loading.
That a designer could design through front loading still does not resolve the issue of who/what the designer really is. Did multi-cellularity arise once all the necessary building blocks were there? And what is the causal direction here? That unicellular life was designed to give rise to multicellular life or that multicellular life arose once unicellular life was sufficiently evolved? Causal direction in issues of evolution can be quite complicated and the additional problem that evolutionary mechanisms can be argued to be designers of front loading complicates the whole picture.
(3.) Thirdly, one can always attribute the origin of such hardware to non-teleological processes, but what a coincidence all this would be. That is, selective pressures shaping protozoan life forms for millions of years just happened to find solutions that would just happen to become essential for metazoan life.
Or we can attribute the origin of such hardware to the processes of evolution namely variation and selection which of course deals with the issue of coincidence in a far more constrained manner than appeal to intelligent designers in that these are actually observed mechanisms. That such mechanisms can be teleological in nature further complicates matter. In fact Mike suggests that metazoan life was expected as an outcome but that is looking back with 20-20 vision. Of course looking back multicellular life seems quite obvious but how is that obviousness explained? By pointing out that basic evolutionary processes can give such an appearance as front loading seems to argue? But where is then the intelligent design relevance of front loading? In other words if evolutionary processes can give result in front loading, then how can we distinguish between intelligently designed and merely designed front loading?
By going beyond the instance of front loading. If there are reasons why we cannot go beyond that instance then we have to conclude 'we don't know', if we can go beyond then we need to establish how plausible evolutionary mechanisms play out against an intelligent designer(s) scenario.
Mike gives an excellent reasoning for why evolutionary pathways may seem front loaded
"The concept of adaptive constraint is another reminder that evolution is not an adaptive free-for-all. Evolution is constrained by the very substrate upon which it is played out. "
By virtue of being constrained, it seems that such evolutionary explanations make for much better explanations.
Another great example by Mike is relevant to my questions and observations about Caporale style front loading
Mike: If coupled with carefully chosen initial proteomes, the potential exists that the first major evolutionary steps subsequent to the originally designed state where rigged such that the mutational bias could untap secondary designs that were front-loaded into the original state.
Front loaded by what? Evolution itself or what? In other words if hydrophobicity itself was selectively advantageous then how does one distinguish between this and intelligently designed front loading?
Can an approach which fully embraces evolutionary mechanisms after the instance of front loading resolve the question about intelligent design? We can all agree that in principle intelligent design is a possibility in this case since we seem to be unable to constrain it in any meaningful manner. At most we can say "we don't know"
[/b]
Charlie stated it well when he argued
quote:
Hmmm... I think that the concept of evolution as a tinkerer (or, more accurately, and much better-sounding, in french, a "bricoleur", is 40 years old, give or take a few). That all existing genes can be eventually tracked back to genes in more ancient organisms is in fact a paradigm of evolution. Of course amphioxus looks like a vertebrate-in-waiting: vertebrates are evolved protochordates. But why aren't mollusks, or insects, anything-in-waiting? That seems to me is the real teleological question (and I'll get back to it).
I think the most telling sentence in John's Morris quote in fact is:
quote:
This, however, is a distorted view. First, many of the basic building blocks and processes found in the metazoans are very widespread among eukaryotes and must have originated much earlier than the first animals. Second, sponges are well adapted, abundant, and diverse. They are highly organized, capable of coordinated response, and despite the relative simplicity of their bodyplan, some sponges demonstrate a radical reorganization concomitant with a shift to carnivory.
That is, while evolution must work with the set of tools it has been given by the contingencies of past history, every step of the way also must follow the contingencies of local adaptation. That is why, of course, sponges and protozoa and bacteria also possess an equally large number of genes that got lost, or were never co-opted into novel and different forms and families in later organisms. That is why eukaryotes had to "reinvent" the flagellum, different species of arctic fish used the same basic strategy to generate anti-freeze proteins, but all started from very different precursors, and insects, birds and bats all came up with different engineering solutions to generate aerodinamic surfaces useful for flying.
Mike in the same thread stated the following which I believe captures my criticism quite well
quote:
If we are to define 'evolution' as change over time, I think you are right. And if such evolution is a necessary consequence of chemistry, it stands to reason that any designer behind life would know about this. Thus, we might very well expect any designer to factor this into the design. Ironically, this would mean that any good designer would make good use of evolution and many later on would view this evolution to mean the original design/designer was superfluous.
Such an approach seems to make intelligent design equivalent to non intelligent design and thus the front loading cannot in itself with additional evidence be used to argue in favor of intelligent design. At most we can argue that 'we don't really know if the initial condition was designed by nature or by intelligence'.
[ 08. December 2003, 00:49: Message edited by: Pim van Meurs ]
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 08. December 2003 21:02
quote:
The assertions are as follows: science requires an independent base of knowledge about the designers to infer design. This is then a limitation of science.
I think that it is rationality itself, combined with a knowledge of the logic of mathematics, that provides this result. Not something that is a special limitation of "science", appart from rational thought itself.
If you think that designers, without independent base knowledge of their probabilities of action through some knowledge of their processes, provide high prior joint probability of producing an effect that is observed, then rationality and mathematics (based on that assumption) can provide a well-supported (by the assumption and mathematicics) conclusion of designer action. However if the prior joint probability of producing the effect observed is low (as compared to other processes joint probability), then mathematics and rationality does not support an inference of designer action.
I have no problem with people asserting that they believe in high joint probability of such cause, even in the absence of any evidence even for the existence of the designers being posited. I only object to claims that evidence and rationality points to that conclusion in absence of the assumptions that must be present for the mathematical conclusion to occur.
This is not a limitation on "science", it is a limitation based on rationality and mathematical logic. Of course one may wish to dispute the value of those disciplines.
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 08. December 2003 22:47
Ged:
quote:
If you think that designers, without independent base knowledge of their probabilities of action through some knowledge of their processes, provide high prior joint probability of producing an effect that is observed, then rationality and mathematics (based on that assumption) can provide a well-supported (by the assumption and mathematicics) conclusion of designer action. However if the prior joint probability of producing the effect observed is low (as compared to other processes joint probability), then mathematics and rationality does not support an inference of designer action.
This is only if you assume that designer action is as a natural law, and therefore, works predictably. This is what is irrational, in my opinion.
[ 08. December 2003, 22:48: Message edited by: Nelson-Alonso ]
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 08. December 2003 23:44
quote:
This is only if you assume that designer action is as a natural law, and therefore, works predictably. This is what is irrational, in my opinion.
But Alonso, you will be surprised to know that I agree for the most part. To assume one can calculate or estimate "designer" probability with narrow bounded probability estimates would indeed be irrational. This is even true in trying to estimate natural cause events with narrow probability bounds--because the information to make probability known within narrow probability bounds is really difficult to obtain. This is a basic problem with concepts like the "explanatory filter", that they are based on probabilities that are very difficult to obtain.
But we note that the probabilities used in the "explanatory filter" are of very widely varying values. Such values of 10^-150 are used, and if they differ, such as 10^-300, we hardly notice that difference. Such calculations can tolerate estimates with very wide bounds on their accuracy--if for example we say only that the probability is below a certain value but it could be many orders of magnitude below that value (or likewise above a value).
In fact those probabilities are not even a constant value of nature, rather are dependent on one's knowledge. If one knew that there was a plan for a murder, one would judge that the probability of murder was higher than if one was not aware of the plan. Yet that event will unfold as it will, unchanged by one's knowledge of the intent of others. Probabilities of events are highly dependent upon the calculations of particulars that are situationally dependent.
If I roll a fair die, and then load a second "loaded" die to match that roll, and then roll the loaded die, with my particular knowledge of the loading value as N I thus know that the probability of rolling value N is very high, and the other values is very low. (A loaded die can still give other values, just distributed unfairly.) But from outside that process, the steps involved still produce a result that is equally distributed. The probability of the loaded die, in that sequence, is dependent upon knowledge of constraints, is "conditional" upon the problem description.
But that is really not at issue here. Because to claim that such probabilities cannot be estimated make the "explanatory filter" or even less numerically based inferences out to be irrational in that case.
My point is not to discuss how we obtain these probabilities--rather I am discussing what we do with them when we obtain them. And we obtain some sort of probability bound even when we speak in general terms.
Consider the probability of a murder occurring. Now if there is no one in the area in question at all, no potential perpetrator, and no potential victim, what is the probability of such an event occurring in the area in question? It must be considered to be very low, virtually zero. It is low because physical constraints (such as lack of a possible victim) reduce the possibility of such an event. This is not 'irrational' to note such physical connections as aspects of reality.
But even beyond this, I am making statements based on any view of what the probability is. Since the explanatory filter and other methods used by ID enthusiasts are based either on probabilty or at least claims of extreme unlikelihood of events, then the ID enthusiasts have introduced probabiltiy into the picture. Assuming that we are not dismissing the ID enthusiasts out of hand for this, then we may consider what the results are in the cases in which we make further assumptions on the probability of other aspects--such as the joint probabilty of the "designer" acting to cause the event.
Alonso, you may choose to claim that making an assumption about the probability of an action occurring canot be assumed, and the case evaluated based on that assumption, if you wish. Without allowing such assumptions to be made, please describe the rationality of the "explanatory filter".
[ 08. December 2003, 23:51: Message edited by: gedanken ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 09. December 2003 00:23
Ged:
quote:
This is even true in trying to estimate natural cause events with narrow probability bounds--because the information to make probability known within narrow probability bounds is really difficult to obtain. This is a basic problem with concepts like the "explanatory filter", that they are based on probabilities that are very difficult to obtain.
Actually, it isn't a problem for non-design processes because they do act in a law-like fashion, with frequencies and probability distributions that we can speak of. Not so for intelligent agents.
You cannot treat a design hypothesis as a chance hypothesis no matter the bound.
I wanted to point out some other points that you bring up.
Ged:
quote:
But Alonso, you will be surprised to know that I agree for the most part. To assume one can calculate or estimate "designer" probability with narrow bounded probability estimates would indeed be irrational.
My comment had nothing to do with "narrow bounded estimates".
Ged:
quote:
But we note that the probabilities used in the "explanatory filter" are of very widely varying values. Such values of 10^-150 are used, and if they differ, such as 10^-300, we hardly notice that difference. Such calculations can tolerate estimates with very wide bounds on their accuracy--if for example we say only that the probability is below a certain value but it could be many orders of magnitude below that value (or likewise above a value).
This also has nothing to do with anything I said. But I'll comment on it anyway. If something is 10^-300 probability, then that can be said to be higher than 10^-1000 and thus much less complex. Just thought you'd like to know that.
Ged:
quote:
In fact those probabilities are not even a constant value of nature, rather are dependent on one's knowledge.
Science is not about absolute knowledge.
Ged
quote:
If one knew that there was a plan for a murder, one would judge that the probability of murder was higher than if one was not aware of the plan. Yet that event will unfold as it will, unchanged by one's knowledge of the intent of others. Probabilities of events are highly dependent upon the calculations of particulars that are situationally dependent.
Here you come back to the Bayesian approach. Oddly enough Ged agrees with Dembski that it is not sufficient:
quote:
That is why Dr. Dembski is correct in NLF page 103, where he shows that simply finding the "best explanation" in a purely Bayesean sense is not sufficient.
but then a few posts later jumps right back into that bandwagon.
Ged
quote:
If I roll a fair die, and then load a second "loaded" die to match that roll, and then roll the loaded die, with my particular knowledge of the loading value as N I thus know that the probability of rolling value N is very high, and the other values is very low. (A loaded die can still give other values, just distributed unfairly.) But from outside that process, the steps involved still produce a result that is equally distributed. The probability of the loaded die, in that sequence, is dependent upon knowledge of constraints, is "conditional" upon the problem description.
Ok, but now connect this in any relevant way to our discussion.
Ged
quote:
But that is really not at issue here. Because to claim that such probabilities cannot be estimated make the "explanatory filter" or even less numerically based inferences out to be irrational in that case.
Your particular knowledge of the loading value as N (or more generally pointing out this sequence of events) has nothing to do with any point discussed in this thread, much less the EF.
Ged:
quote:
Consider the probability of a murder occurring.
Now if there is no one in the area in question at all, no potential perpetrator, and no potential victim, what is the probability of such an event occurring in the area in question? It must be considered to be very low, virtually zero. It is low because physical constraints (such as lack of a possible victim) reduce the possibility of such an event. This is not 'irrational' to note such physical connections as aspects of reality.
This is getting a little relevant but not quite. The problem with ID and ID related issues is that we have a "victim". Or more accurately, we have a corpse. What we have to do now is look at the properties of the system (the corpse) and see if there are signs of murder, if it was a natural cause, or accidental. We can use Fisher's approach (eliminating natural and accidental causes) in order to get at design.
Ged:
quote:
Assuming that we are not dismissing the ID enthusiasts out of hand for this, then we may consider what the results are in the cases in which we make further assumptions on the probability of other aspects--such as the joint probabilty of the "designer" acting to cause the event.
Alonso, you may choose to claim that making an assumption about the probability of an action occurring canot be assumed, and the case evaluated based on that assumption, if you wish. Without allowing such assumptions to be made, please describe the rationality of
the "explanatory filter".
In the case general ID, the filter eliminates chance and regularity and simply leaves questions about the mechanism of design, who is the designer, etc as a seperate question. However, it doesn't make sense to talk of "probability of design acts" because they simply aren't natural processes.
[ 09. December 2003, 00:59: Message edited by: Nelson-Alonso ]
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 09. December 2003 00:41
Nelson: Actually, it isn't a problem for non-design processes because they do act in a law-like fashion, with frequencies and probability distributions that we can speak of. Not so for intelligent agents.
Tell that to the poll takers :-) In all seriousness, while in principle such frequencies and probabilities may seem to be non-problematic, applying these concepts to real life examples seem to be quite complicated. Take the flagellum for example, how does one calculate the relevant probabilities in a non strawman manner?
Nelson: You cannot treat a design hypothesis as a chance hypothesis no matter the bound. This pretty much addresses the rest of your post.
Now does it? Unless design hypotheses can be bounded in some form or manner, inferring intelligent design may be quite complicated. Take for instance criminology where motives, likelihoods, possibilities are used to rule out or rule in potential witnesses, including profiling.
As Gedanken carefully argues, the problems of accurately calculating probabilities, for design and intelligent design alike may be too complicated for a design inference based on the explanatory filter to succesfully work.
Gedanken continues showing how in criminology, probability bounds are used to eliminate or include potential 'designers'.
If you want to argue that probability calculations are inherently unreliable then there seems to be a problem.
It all comes down to what Gedanken argues to be
quote:
Assuming that we are not dismissing the ID enthusiasts out of hand for this, then we may consider what the results are in the cases in which we make further assumptions on the probability of other aspects--such as the joint probabilty of the "designer" acting to cause the event.
Given Alonso's response I am curious as to his answer to Gedanken's question
quote:
Alonso, you may choose to claim that making an assumption about the probability of an action occurring canot be assumed, and the case evaluated based on that assumption, if you wish. Without allowing such assumptions to be made, please describe the rationality of the "explanatory filter".
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 09. December 2003 00:47
I must complement Alonso for arguing this to the bitter end, rather than changing the subject. This is the best that ID has to offer on this. Since the "explanatory filter" does not withstand the harsh glare of mathematical analysis, debate must make such analysis seem irrational itself.
(Acutally I must amend this, as Alonso added in edit a bunch of other aspects. I shall not attempt to answer his point by point quotes.)
But of course I am not (once again) claiming to analyze the designer in terms of a claim of mechanistic process. Rather I am simply stating in the abstract an analysis of what happens if designer actions happen with a given frequency, taken as probability.
ID analyses such as the "explanatory filter" are often given in terms of a flow chart of decisions based on probability of events, an abstract decision tree based on abstract values of those probabilities. And then the "explanatory filter" algorithm is expected to be applied to concrete values of those probabilities. It is an abstract case analysis, to which one gives concrete values and then a result pops out of the algorithm steps.
I am simply performing such a case analysis of the "explanatory filter" itself, giving what happens under these abstract cases. I assume that lack of knowledge of the probability of a law-like event in a particular case does not invalidate the meaning of the "explanatory filter" as an abstract procedure. No more should lack of knowledge of the probability of the designer's actions in a particular case invalidate an abstract analysis of the EF itself.
I'm perfectly content that the readers should know that the best defense of the EF must rest on claiming that the making a case analysis assumption that a designer could act in a certain frequency (taken as probability) must be denied.
(I hope that those who collect police statistics are reading, and now know that their efforts are irrational!)
--
Just wanted to add a case for consideration. Consider a region in which there are no designers, no designing processes. What is the probability of "design" occurring to produce a particular class of result?
Now of course one may ask: How can one know that there are no designers able to act? To that I ask, how does one know that there are no unknown natural processes causing the event (processes that we didn't think of during our present level of analysis)? How does this differ? But surely if one says -- as an assumption -- that the probability of event the presence of designers is below 10^-150, then the probability of "design" of a particular event must be considered to be below that value, given the assumption.
[ 09. December 2003, 01:03: Message edited by: gedanken ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 09. December 2003 00:58
Ged
quote:
But of course I am not (once again) claiming to analyze the designer in terms of a claim of mechanistic process. Rather I am simply stating in the abstract an analysis of what happens if designer actions happen with a given frequency, taken as probability.
Thats like saying "I didn't say I want to kiss you, I said I want to touch my lips to your lips".
Ged:
quote:
(I hope that those who collect police statistics are reading, and now know that their efforts are irrational!)
This is after-the-fact information and has nothing to do with , as you call it, "probability of space ships".
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 09. December 2003 01:05
quote:
quote:
But of course I am not (once again) claiming to analyze the designer in terms of a claim of mechanistic process. Rather I am simply stating in the abstract an analysis of what happens if designer actions happen with a given frequency, taken as probability.
Thats like saying "I didn't say I want to kiss you, I said I want to touch my lips to your lips".
So Alonso, you are saying that police statisticians and poll takers are (mistakenly) making "mechanistic" assumptions about designers of opinions and crimes?
And as to "after the fact" taking of statistics--are you denying that statistics is based on probability? That would be interesting, how do they ever get all those results to analyze reliability of statistical procedures, error bars, etc? So they can't actually make any probabiltiy assumptions in analyzing gathering of statistical information--if that statistical information happens to be a matter of "design" issues such as crime or opinion polling. Yes interesting...
Indeed, if this is the best that ID can offer, I do find it interesting.
[ 09. December 2003, 01:11: Message edited by: gedanken ]
IP: Logged
|
|
|
Printer-friendly view of this topic