|
Author
|
Topic: Nature Refutes ID?: The Evolutionary Origin of Complex Features
|
Pim van Meurs
Member
Member # 541
|
posted 26. May 2003 15:30
Y2Z: The behaviour of each of the EQU programs using the registers is already quite clear. They are just different partitions of the problem subspaces with the application of an existing NAND on the content of the registers. No more, no less. Without the logical meaning of EQU, the programs are as meaningless as random strings.
Specified complexity indeed I would argue. Without a specification they are 'as meaningless' as random strings.
Does the same not apply to the flagellum for instance? Without the specification "propulsion" the flagellum would seem to be as meaningless, generated by a sequence of nucleotides which are part of the full sequence space.
The question is and seems to have been resolved in the affirmative by the work of Lenski et al here: Can RM&NS explain the arrival of such complex specified systems?
IP: Logged
|
|
YZ2
Member
Member # 91
|
posted 26. May 2003 15:50
Thanks, Pim. I think you have hit on the major point. Just as in Lenski's study, EQU actually has TWO parents. It arrives through RM/NS process (at least for this case), but it is also 'selected' by the selection function. Ignoring one is like claiming this is borned through a cloned procedure (and the complexity will deem there is no evidence for it)! I hope you like the analogy.
[ 26. May 2003, 15:58: Message edited by: YZ2 ]
IP: Logged
|
|
YZ2
Member
Member # 91
|
posted 26. May 2003 17:11
I probably need to qualify the previous statement, so that I will not be misunderstood.
I mention already that it can happen does not imply that it has happened. This experiment has shown that RM/NS + selection function can generate IC and ID. However, whether a past event has occurred (such as complexity in life has evolved) can only be evaluated by a legal argument in terms of its doubt or belief. It is still uncertain, but at least its belief can be evaluated. The reason: it is possible does not imply that it has occurred. In this sense, this experiment is quite irrelevant in determining whether complexity in life has evolved in the past.
[ 26. May 2003, 17:20: Message edited by: YZ2 ]
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 26. May 2003 18:14
YZ2 wrote quote:
One more point:
Nobody claims that the observer defines design in the Beethoven's fifth experiment. The musician defines it (not the child). The observer estimates (or detects) it.
But that the observer must estimate it subjectively is precisely the problem. An operational definition does not depend on the subjective state of knowledge of an observer as a detection device. The knockout criterion defines a procedure that any observer can use to determine ICness; the subjective state of prior knowledge of Beethoven's Fifth is not such a procedure.
YZ2 further wrote quote:
I mention already that it can happen does not imply that it has happened. This experiment has shown that RM/NS + selection function can generate IC and ID. However, whether a past event has occurred (such as complexity in life has evolved) can only be evaluated by a legal argument in terms of its doubt or belief. It is still uncertain, but at least its belief can be evaluated. The reason: it is possible does not imply that it has occurred. In this sense, this experiment is quite irrelevant in determining whether complexity in life has evolved in the past.
The experiment showed that IC structures can evolve; it does not show that ID was "generated" in the sense of 'intelligent design of an irreducibly complex structure that performs a specified function.' To repeat once again, the programs that evolved to perform EQU were not themselves designed.
Further, the claim by Behe and Dembski and their supporters has been that IC structures cannot evolve, and therefore it is necessary to invoke an intelligent agency to account for them. The experiment shows that IC structures can evolve via the mechanisms identified by evolutionary theory. As a consequence, the argument that complexity in life could not have evolved because biological structures have the property of unevolvable ICness fails. IDists cannot now argue that complex biological structures could not have evolved because they are irreducibly complex; that argument has lost all force. Thus it is completely relevant to the question of whether complexity in life evolved in the past by showing that the IDist argument (that irreducibly complex structures and processes cannot be produced by evolutionary mechanisms) is empirically false.
That it is possible does not imply that it occurred; but the demonstration that it is possible implies with necessity that the claim that it is impossible is false. And the claim of impossibility is the argument ID has made with respect to irreducible complexity. That's the core of the so-called "design inference."
RBH
[ 26. May 2003, 18:24: Message edited by: RBH ]
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 26. May 2003 19:06
Y2Z, Mutation/variation AND selection are intricately connected. But you have to realize that it is the ID crowd who has claimed that IC systems could not evolve through (Darwian) pathways. Now that ICnesss itself is NOT a reliable indicator of design, as was previously argued, ID has to look for another approach.
IP: Logged
|
|
Micah Sparacio
Member
Member # 6
|
posted 27. May 2003 09:05
quote:
To repeat once again, the programs that evolved to perform EQU were not themselves designed.
Sure they were. You've just got a narrow view of design. Carry your logic a bit further, and we could say that no program is designed, and design thus loses its meaning altogether (which, I know, is what some have been arguing all along on various forums). When I program, for example, I don't exert any control over the individual pulses of electronic current in the CPU. There are many things that I am not in direct control of as I program. However, it seems that this is precisely what RBH is requiring for something to be designed.
All of Mike Gene's refreshing ideas...forgotten in a flash. We should all be able to agree that design can take place on many levels, and it need not be the sort of micromanagement design that design critics like to strawman (yersina comes to mind here).
[ 27. May 2003, 09:07: Message edited by: Micah Sparacio ]
IP: Logged
|
|
YZ2
Member
Member # 91
|
posted 27. May 2003 09:16
Since the other points have been answered, I only need to address whether the observer must require subjective knowledge to detect design. I only need to point out that there are many active fields that have been very fruitful in doing research in these areas, including the traditonal statistics (which uses very little "subjective" knowledge other than modeling), machine learning and knowledge representation (which uses "more" subjective knowledge), pattern recognition and information theory. Unless you think these are not science, then I cannot really say anything, but there will be many scientists who will disagree with you.
[ 27. May 2003, 09:55: Message edited by: YZ2 ]
IP: Logged
|
|
YZ2
Member
Member # 91
|
posted 27. May 2003 09:44
If you have difficulty understanding why EQU has both parents. My Beethoven’s fifth experiment is a better illustration (and please do not call this by any other name). When the light sequence evolves (and I love this term) from the child’s play to the musician’s touch, you can say that the light sequence is created by BOTH the child and the musician. I have no objection to this interpretation. After all, without the child’s action, there will be no Beethoven’s fifth light sequence. Now clearly the Beethoven’s fifth light sequence is designed, whether the observer detects this or not. There is an inherent meaning of IC and ID, whether you identify them or not. On the other hand, it is also possible by chance, the child’s action creates a light sequence that is similar to the first bar of ‘Mary has a little lamb’ and the observer has also identified it. This small sequence is not designed and not intended, but it appears to have the characteristics of design. This is actually a case of false positive of design.
[ 27. May 2003, 09:47: Message edited by: YZ2 ]
IP: Logged
|
|
GP
Member
Member # 570
|
posted 27. May 2003 10:31
I finally had some time to read this thread. And I shall jump right into the fray with the most recent response. Micah, you wrote: quote:
You've just got a narrow view of design. Carry your logic a bit further, and we could say that no program is designed, and design thus loses its meaning altogether (which, I know, is what some have been arguing all along on various forums).
Actually, I agree with you here, Micah, in part. But I also happen to agree with RBH when he writes: the programs that evolved to perform EQU were not themselves designed.. As I struggled to reconcile my viewpoints, I realized that the underlying problem for me is that the notion of design is so ambiguous that it readily permits any interpretation, and if the circumstances permit, allows anyone to score debate points. The mistake I believe the Darwinians are making in this thread is by playing according to the rules of the IDists (by using their definitions and so forth) -- when IDists have not published a tested and standardized rule set. Upon submitting paper for peer-review, one is often asked whether the conclusions are justified by the data. Here, the authors (Lenski et al.) do NOT talk about the implications with respect to design (and rightly so), not only because it makes for poor science to introduce such philsophically ambiguous concepts as intelligent design and front-loading into the publication, but also because the evidentiary support is most effective towards a Darwinian paradigm than a design paradigm.
Since my joining this forum, I have had on occasion asked several IDists to define what it means for something to be designed. I believe the closest came when Paul Nelson[*] wrote me: quote:
For design (as a real cause, and scientific explanation) to have any empirical content, it must have a contrast class -- namely, natural causes. Now, if one requires the logically impossible, i.e., that we exhaust the universe of possible natural causes, one cannot infer design.
Even in the critical thread about Dembski's CSI concept, I asked once again what it means to be not designed. And there, Nelson Alonso[+] tells me: quote:
Indeed, there are many false negatives in Biology that provide fruitful research prospects for the design inference (i.e. sub-optimal design). A true negative can also be given, as for example, the hemoglobin case, where the specification is quite small and natural selection can indeed select functional intermediates.
As far as I can see, there is only one common theme running in these two quotations: design is not natural, not evolved. So y'see, RBH isn't the only person with a narrow view of design here when he's in good company with the likes of the Nelsons, nor should he be faulted for picking up on the design notion du jour on an IDist site. Or to bring the point home, if we carry Micah's logic all the way to the other end of the spectrum, everything is designed -- for that is the broadest notion of design possible. And why not? The real shame of it is that design as a concept (as evinced by Nelson) is defined by its antithesis (supposedly natural evolution).
This all brings me back to the time when I asked Nelson Alonso[-] [towards the end of a thread on Brainstorms entitled "Evolving Inventions"] for an experimental setup to test the IC argument (not necessarily the IC definition, as I understand it). The reason I find that old thread relevant is because it gives us a sense of how an IDist would approach the experimental support. Suppose I grant Micah's assertion that EQU is according to his design-sense "designed" because the experimental conditions favored (or rather expected) a particular outcome. How then does one go about scientifically testing whether an IC structure can in fact be evolved? If every experimental condition has on some level a designed component, I am interested in hearing how one can design an experiment that controls for the "front-loading" aspect of design. Do we really want to follow Alonso's suggestion (via Behe, apparently) in putting bacteria in a jar and see if we can observe some miraculous event can occur within our lifetime? Well, Alonso [-] apparently thinks that even his own experimental setup won't impress him: quote:
What would happen if we removed that and reverted them to their supposed ancestral state? Thats what the experiment seeks to look for. If bacteria do this by reengineering their genome in a complex way, that still wouldn't be Darwinian evolution.
. Maybe another ID proponent has a different opinion.
To reiterate, I honsetly would like to hear some alternatives to the Avida experiments. And I'd love it more if IDists actually conducted them, instead of proposing them with caveats to the effect that they are unqualified or not funded, etc. to conduct the experiments.
I'll close with the observation that ID proponents should find no comfort in reading that the Avida experiment does not explicitly state its implications on the popular ID concepts of IC or CSI. I believe it a mistake to think that the mainstream scientific literature has any vested interest in addressing ID arguments when they could be supporting Darwinian ones.
Signing off.
GP
[*]http://www.iscid.org/boards/ubb-get_topic-f-6-t-000288-p-2.html
[+]http://www.iscid.org/boards/ubb-get_topic-f-6-t-000353-p-3.html
[-]http://www.iscid.org/boards/ubb-get_topic-f-6-t-000287-p-13.html
[ 27. May 2003, 10:53: Message edited by: GP ]
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 27. May 2003 11:09
GP wrote quote:
As far as I can see, there is only one common theme running in these two quotations: design is not natural, not evolved. So y'see, RBH isn't the only person with a narrow view of design here when he's in good company with the likes of the Nelsons, nor should he be faulted for picking up on the design notion du jour on an IDist site.
In fact, my own conception of "design" is not coterminous with what the IDist notions of irreducible complexity and specified complexity would imply. Read Herbert Simon's The Sciences of the Artificial for a pretty good overview, though it's a little dated now.
One accepts the IDists' definition for the sake of this discussion in order to ascertain whether a central empirical claim of two major ID proponents, Behe and Dembski, holds water. In order to evaluate that empirical claim one must use the terms of the claimants. And the conclusion I reach is that evaluated in its own terms, the ID claim that complex structures cannot evolve (either in principle or with cosmically low probability) is false.
The questions of ambiguity and subjectivity are now new. Every extended discussion of the IDist claim of the impossibility of the evolution of complex (in one or another sense) structures sooner or later encounters that subjectivity, that lack of operational definitions. See the discussion of "well matched" above.
I agree that "design" as it is used in these discussions is a slippery concept. And "intelligent design" is not much better. In fact, as I read ID proponents, I have come to believe that "intelligent" and "design" are superfluous: the core issue is intentionality. Yet that rarely gets directly discussed. The view is apparently that intelligent agents must have intentions, so if intelligent agency can be implicated in 'natural' events, then intentionality is established a fortiori.
RBH
IP: Logged
|
|
Argon
Member
Member # 276
|
posted 27. May 2003 11:34
Micah writes:
quote:
All of Mike Gene's refreshing ideas...forgotten in a flash. We should all be able to agree that design can take place on many levels, and it need not be the sort of micromanagement design that design critics like to strawman (yersina comes to mind here).
I would agree that design can take place on many levels.
Two words come to mind: "Proximate" (direct intervention immediately prior to the emergence of a designed system) and "distal" (i.e. front-loading).
With regard to IC systems, please note that it was proposed by Behe and Dembski that these systems could not arise via natural mechanisms. And by natural mechanisms, I don't mean starting with a glass of water and some carbon, but from pre-existing organisms. The argument was "How do you get from organism-X without the IC system, A1, to organism-Y with the A1 system?" Thus, the enactment of design leading to the emergence of an IC system was initially described as requiring a proximate intervention, not a distal one. Otherwise, how does one reconcile the preoccupation with things like the blood clotting cascade, the immune response and the bacterial flagellum? Why spend time discussing the difficulty of evolving a clotting system from an earlier ancestor if that parent already had the "seeds" of a potential clotting system pre-loaded within? This is not to say that front-loading is not a possible design mechanism, but it does complicate the determination of what things were designed. It certainly eliminates Behe's original concept of ICness as an indicator of design because "front-loaded IC" is an oxymoron.
IP: Logged
|
|
YZ2
Member
Member # 91
|
posted 27. May 2003 12:03
GP,
Thanks for your comments and I do understand the problem that you have raised. I do not consider myself an IDist (at least not the traditional sense), but I try to understand the problems and clarify them.
I think as a scientific paradigm, ID theory has raised an important issue, i.e. that complexity in life cannot be fully explained from the evolutionary standpoint. There is another view from the design standpoint that can explain fruitfully the biological nature. That is, in addition to the explanation from the fitness function, there is an effect from the selection function. This is a fully legitimate scientific concern, and many fruitful research can come out from it. Ignore this is a major hindrance to the advancement of science.
Now design can be caused by natural or supernatural means. I believe this is an important question to ask, but this can be formulated as a secondary issue after the issue of design (natural and/or supernatural) is addressed. Personally, I do not even know whether this can be completely settled using scientific means. I am not saying that methods cannot be formulated to address, at least some of this issue about supernatural design. However, without developing research to address detection of design (natural/supernatural) in biology or suppressing it because of philosophical or political correctness, everybody, scientist and non-scientist will be hampered as a result.
Now with respect to the effectiveness of the proposed ID methods of detecting design. My understanding is that if we accept that there can be design in biological nature (natural/supernatural) and is important to detect them, then it is not an issue whether our current method is sufficent enough, it can be refined if we allow the research to proceed. That happens historically to the evolutionary theory, and perhaps still does. Afterall, design in nature ultimately is to be ESTIMATED only, and so is the evolutionary mechanism. You do not know completely certain whether a phenomenon is designed (as illustrated in my Beethoven's fifth experiment). But the knowledge gained from it can be very valuable.
[ 29. May 2003, 11:09: Message edited by: YZ2 ]
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 28. May 2003 23:48
Y2Z: I think as a scientific paradigm, ID theory has raised an important issue, i.e. that complexity in life cannot be fully explained from the evolutionary standpoint.
ID has questioned this but I do not believe that ID so far has made a solid case for this. In fact the recent publications suggest that complexity in life may indeed be explainable from an evolutionary standpoint.
Y2Z: Afterall, design in nature ultimately is to be ESTIMATED only, and so is the evolutionary mechanism.
Not really. Design in nature is infered differently than evolutionary mechanisms. The latter one is based on direct observations namely the existence of variation and natural selection, the former, well... The moderator would object to me finishing my sentence so I will leave it on a positive note namely that ID does not use the same approach to infering design.
Could ID contributions be valuable? Of course, after all teleology in some form or manner has been quite helpful in formulating hypotheses about nature. But it is important that we understand the differences between internal and external teleology and not conflate them
IP: Logged
|
|
YZ2
Member
Member # 91
|
posted 29. May 2003 09:58
Pim,
Indeed you are probably correct in many ways. These are my thinking:
1) I think whether complexity in life has autonomously arised in the past can be evaluated based on the best judgement on the evidence, rather than based on a theoretical formulation from it. It is also more important to understand
how actually it occurs. I am certainly not an expert on this topic. And I am looking forward to experts who can shed more lights on this.
2) In terms of theoretical formulation, ID with an evolutionary mechanism can be a potentially richer theoretical framework for studying biology. If this is true, then this is a good news. But it needs to be developed further.
3)Between ID and the Darwinian approach, they are (or can be) different. My understanding is that the Darwinian approach is more a model-first approach,but the ID approach is more a data-first approach, generating a design model from the data and then inferring properties from the model and/or the data, or something similar. It is a little complicated, similar to a recurrent definition. It is not new in engineering. For example in pattern classification, there are the unsupervised and supervised methods.Often you use an unsupervised method to generate a model, and then use the generated model for supervised classification. If the generated model is not sufficent, then you perform another iteration if possible. The whole method is still scientific. In practice of course it can be more difficult. There are obviously a lot of theoretical and application works that need to be done. My wish is that many traditional evolutionists can join in developing the ideas.
I hope this helps.
[ 29. May 2003, 12:11: Message edited by: YZ2 ]
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 29. May 2003 12:37
Y2Z: 2) In terms of theoretical formulation, ID with an evolutionary mechanism can be a potentially richer theoretical framework for studying biology. If this is true, then this is a good news. But it needs to be developed further.
ID with an evolutionary mechanism sounds almost self-contradictory. In order to infer design one has to show that it is specified complex or "...highly improbable with respect to all causal mechanisms currently known"
source
quote:
Bottom line: Darwinism has a burden of proof that intelligent design does not have. Darwinism is a theory of process and therefore needs to provide convincing evidence that the processes it describes are able to bear the weight placed on them. That weight is considerable -- indeed, no less than the whole of biological complexity and diversity. Intelligent design by contrast has a different burden. As a theory of creative innovation, its burden is to show where creative innovations first emerge and then trace their causal antecedents and consequents. Darwinism and intelligent design therefore face fundamentally different tasks, though there will be considerable overlap in their practical outworking (intelligent design, for instance, does not require that every aspect of biology be designed and is fully capable of assimilating the Darwinian mechanism).
Source
IP: Logged
|
|
|
Printer-friendly view of this topic
