|
Author
|
Topic: The Wings of Eagles
|
Frances
Member
Member # 169
|
posted 11. August 2002 13:29
Warren:
quote:
The biological reality that I have been pointing out is that 1)RM&NS has not, and apparently can not be reconciled to the known facts regarding the redesign of eagles wings, and 2)a ‘directed’ design model is consistent with known facts and is explainable in terms of likely physical mechanisms. Design theory, in the example being discussed, appears to describe biological reality. Neo-Darwinian genetic theory, it appears does not and can not describe biological reality.
I find your claim 1) to be fascinating. Perhaps at most your claim could be that it has not yet been reconciled but the claim that it cannot be reconciled requires a bit more supporting evidence. 'Design theory' indeed can describe many things, in fact I would argue that there are no limitations to what design theory can 'explain'. It's this unlimited level of explanatory power that makes design theory of such limited value. Perhaps your claim that neo-Darwinian theory does not descibe biological reality could benefit from some supporting evidence. In our discussions on this topic I feel that such evidence has been severely lacking.
What am I missing here?
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 11. August 2002 16:05
quote:
You make claims that "your biology has no relationship with the real world". You make such comments with the suggestion you have at least an elementary knowledge of science and biology. You should therefore be able to provide some substantiation for your comment.
Well, for instance the idea that wings length is determined by one gene with 10 alleles determining progressively increasing length is preposterous - as I discussed above, any minimally acceptable model of wing evolution would need to be multigenic. So is the idea that "directional mutation" (of which no convincing empirical evidence exists, while there is solid evidence of the contrary) would be required for gradual phenotypic change, and that incredibly high mutation rates would also be necessary (they aren't, see below). (All these statements just from one paragraph, below)
quote:
There are a large number of neo-Darwinian genetic models that might fit the pattern of change described above. The simplest would be a gene H which defines length and where H has different alleles associated with length L0, L1, ..,L10. As should be obvious, there are a lot of problems with this or any known model of gradual genetic change. The two most obvious are 1)how does the organism ‘direct mutations’ or that L2 produces the L3 mutation and not the L9 mutation and 2)how do you get mutation rates high enough to account for the rates of change that can be produced in selective breeding.
Just for fun, however, consider the following (also completely useless, though not as preposterous) model. Let's say the difference between long- and short-winged eagles' wings is determined by 10 genes (A,B,...X), each contributing 10% of the difference. Each gene can have, let's say, 10,000 non-neutral mutants (i.e. mutants that are able to give a phenotypic change). Of these, 9,999 have a negative impact on fitness (crippled wings and such), and 1/10,000 is the one (allele XL) that makes the slightly longer wings. The small length increase (10% of the final difference) brings about a tiny increase in fitness (better flying), let's say 1%. Let's also say that the initial eagle population is 10^6, and that no positive effect alleles are present at the beginning.
If eagles' mutation rates are similar to those observed in mammals, each individual carries between 0.1 and 1 new phenotypically detectable gene mutations. That is, assuming 50,000 genes in eagles (as many as we do), each gene (including X) displays between 2 and 20 new mutants in each generation in the 10^6 population (let's say 10). This means that a new mutant XL appears every 1,000 generations among the eagles (remember, 1/10,000 X mutants are XL). Because its selective advantage is 1%, its chances of fixation are 2%, that is 1 every 50,000 generations allele XL will become fixed by its own contribution to fitness. The time required for fixation is, as I discussed in my previous post, a few thousand generations (in fact, a little less than 3,000, according to Graur's, I just checked). So, on average, every 53,000 generations, allele XL will appear de novo and completely take over the eagle population. In eagle generations, that's 200-250,000 years.
There are some possible corrections to the model: the ratios of positive to negative mutants could be lower, the mutation rate also lower (unlikely), the fitness advantage or the effective population size smaller. On the other hand, this is balanced by the fact that there are 10 genes evolving independently (not successively as you claim) - a 10-fold acceleration -, and that carriers of 2 or more favorable mutants (which would become common as frequencies rise), may have a bigger than additive fitness advantage compared to single carriers (further acceleration). Also, we are assuming the eagle population started with no positive favorable mutant genes, but in fact in a 10^6 population, a large amount of variability would already pre-exist solely on the basis of genetic drift (that is, no selection) - just think how much variation there is among 10^6 humans as far as arm length is concerned, presumably with no selective advantage either way.
So, on or off by an order of magnitude, that's pretty impressive, for a system that "could not explain or simulate [gradual] evolutionary changes in eagles wings", isn't it? In fact, and this is hard for many non-biologists to understand, darwinian mechanisms are surprisingly efficient.
Now, let me be very clear: my example is totally unrealistic as well, and most importantly it does not teach us anything about wing evolution. It's just something a poor exasperated biologist can come up on a sunday afternoon, rather than take severe punitive action against his screaming children....
Oh, I also like the following:
quote:
In more formal terminology, the analysis of eagles wings suggests that the development or evolution of wings included the development or evolution of some as yet unidentified mechanisms to quickly adapt the shape of the wings to fit environmental conditions.
Of course, by extension, what you are claiming here is that all the genes necessary to make a chihuahua were designed into the wolf's genome with the idea such phenotypic changes will eventually fit the esthetics of 20th century matrons. Or am I missing something?
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 12. August 2002 10:48
Frances,
You raise a very important issue, ‘supporting evidence’. Answering your question, however, is going to require a bit of background. The validity of my statement depends on the formal, ‘hard science’ concept of ‘design theory’. This concept probably needs some explaining.
To begin, the term ‘design’, as used here, refers to a specific logical property exhibited by certain types of systems. I recognize two basic types of design 1)adaptive design (also called simple, elementary, teleological or purposeful design) and 2)biological design(also called intelligent design). Adaptive design is a property or set of properties exhibited biological systems, artificial man-made systems, and some naturally occurring non-biological systems. Biological design is the special, more precise, more complex form of adaptive design exhibited by biological systems. The discussion here only concerns adaptive design. Since biological design includes the adaptive design property, biological systems can be formally analyzed in terms of the adaptive design property.
I introduced a formal definition of the adaptive design property yesterday under a different thread, but the definition is worth repeating.
DEFINTIION OF ADAPTIVE DESIGN(also called simple, elementary, purposeful, or teleological design): - A system, process, or mechanisms is defined as having adaptive design, if 1)the system can take an identifiable number of different forms, 2)some subset of the set possible forms is adaptive or purposeful (increases the likelihood that some goal or function will be achieved), 3)processes or mechanisms exist which are capable of changing the form of the system, and 4) there is over time a greater likelihood the system will be in an adaptive state rather than an non-adaptive or mal-adaptive state.
Note that most traditional concepts of design refer to 1)the state of a system at a point in time and/or 2)a single set of processes for changing a system from a non-adaptive state to an adaptive state. Defining design as an ongoing dynamic process, may be a more complex approach, but, I claim, it turns out to be a far more productive approach. [This dynamic approach also provides a much realistic description of the design and design processes associated with biological systems.]
A ‘design theory’ in the context of this dynamic definition of design is an algorithm or set of algorithms that simulate the processes by which the system being analyzed maintains an adaptive state. It is useful to visualize this type of design theory as a computer program that is continuously performing design or redesign operations in order to maintain an adaptive state.
It is probably not immediately obvious, but this seemingly complex approach makes it possible to apply the rigorous hard science form of the scientific paradigm, the same form used in physics, to the analysis of living systems. Using the design approach, it is possible to precisely define and model the static design of a system at different points in time. This is logically equivalent to the process used in physics of precisely defining topology and the path or movement of objects in the defined topology.
Given the ability to measure and model the path of a design/redesign process, it is then possible to construct models and/or theories which attempt to simulate and predict redesign processes. This brings us to what are arguably two of the central issues in theory construction and testing in the life sciences. Living systems are extremely complex involving large sets of complex causal relationships. The first issue, how do you construct models and theories of systems involving large sets of complex causal relationships. Second, once you have constructed such models or theories, how do you tell if one theory is superior to another.
Design science, the formal scientific analysis of adaptive and biological design, in the form I am describing here, appears to offer answers to both questions. First, it appears that all aspects of life can be analyzed in terms of design. [As proposed here, biological design is the logical property differentiating life from non-life.] Second, and more specifically relevant to the discussion here, the goal or purpose of design is maintaining an adaptive state. The purpose of adaptive states is the continued existence or survival of the system. Different models or theories are evaluated based on their success or effectiveness in promoting survival.
Real world adaptive and biological systems contain noise or inefficiency. They are thus only partially effective at promoting survival. A theory of adaptive or biological change, as with any theory, represents change under ‘ideal conditions’. A model or theory passes a test if the model can match or exceed the performance of the real world system (recognizing the noise or constraints applicable to the real world system(and is falsified if it can not). When comparing two models or theories, one theory is better than another if it is more likely to promote survivability. The wings of eagles example provides a concrete example of how this works.
I proposed a relatively simple model of wing design involving a set of ‘near continuous’ variables such as length and width, where the values of each variable change in incremental steps. Consistent with what is known from selective breeding, the condition is imposed that redesign must involve a series of changes involves one increment at a time. The model of the design/redesign is obviously quite simple, but it is still more accurate and detailed than a simple one step, one mutation model.
For the proposed model of wing design, two different models or theories of the design/redesign process are considered. One redesign process was based on the neo-Darwinian mutate-select model, and one was a directed change model based on the proposed ‘add-delete’ mechanisms. Finally, we can get to your issue of ‘evidence’.
In testing and evaluating scientific models and theories, there are basically two types of evidence-1)consistency with existing facts and 2)valid predictions. Generally, valid unexpected predictions, or predictions that contradict an existing theory or model, are considered more significant than demonstrations that a model or theory is consistent with some known fact. Yesterday I presented three specific predictions generated by the directed design model which directly contradict predictions produced by neo-Darwinian models.
Two of the predictions offered would, I think require some detailed research to confirm or deny. The speed ‘prediction’ is in fact a known phenomena. Multi-step mutate select processes are extremely slow using realistic mutation and selection rates. Darwin’s finches would have been extinct long before they could have evolved based on a multi-step mutation-selection process. If you know math, you should be able to do the demonstration in your head. [Note: This evidence merely shows that for the specific example of redesign of eagles wings, (and probably for most step by step changes in continuous variables) evolutionary change probably involve some type of non-random directed change mechanism.]
I believe I am actually providing hard scientific evidence in support of my assertions. Scientific evidence in relationship to scientific models and theories is not the same as ‘a long list of published articles’. The confusion, I believe, arises in part because ‘design science’ as I am defining it, is significantly different form evolutionary biology as it is currently practiced.
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 12. August 2002 12:46
Warren
quote:
DEFINTIION OF ADAPTIVE DESIGN(also called simple, elementary, purposeful, or teleological design): - A system, process, or mechanisms is defined as having adaptive design, if 1)the system can take an identifiable number of different forms, 2)some subset of the set possible forms is adaptive or purposeful (increases the likelihood that some goal or function will be achieved), 3)processes or mechanisms exist which are capable of changing the form of the system, and 4) there is over time a greater likelihood the system will be in an adaptive state rather than an non-adaptive or mal-adaptive state.
Seems that adaptive design and biological design are indistinguishabele. So if biological design can also be 'adaptive' design according to your definition, what have we gained?
Your claim to be providing 'hard evidence' but all I have seen are 'soft assertions'. It would be beneficial for the discussion if you could support your imho ad hoc 'conclusion' about non-random directed change mechanism. Especially since you start off with a 'formal definition' of adaptive design which seems to include biological 'design'.
Also could you please explain what the algorithm for intelligent design is? I have always been interested in finding out more about the mechanisms of ID.
Also the claim that (intelligent) design will be faster remains unsupported. In fact I could provide you with some examples in which RM&NS seems to have been much more productive in finding solutions than intelligent designers.
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 12. August 2002 15:26
Charlie,
You made a valiant, but obviously unsuccessful attempt to describe how a ‘simple’ 10 step change in size of a single variable might be accounted for using mutate-select genetic concepts. A couple of key elements of your attempt should be pointed out. First, and most obviously, you failed completely to come anywhere close to the redesign speeds demonstrated by selective breeding. The central point of the alternative model I offered is to explain rapid rate of adaptive change that appears to occur in nature that and can be induced under artificial conditions.
In order to develop a mutate-select model to fit step by step changes you are forced to speculate on the existence of very complex genetic coding mechanisms. Mechanisms which would require huge numbers of genes and for which there is absolutely no evidence.
As an aside, I find it difficult somewhat difficult to believe that there isn’t a significant body of knowledge on the mechanisms underlying ‘small incremental changes in the variables controlled by developmental processes’. Such mechanisms would be relatively easy to analyze and of obvious direct interest to genetic theory. The information must either exist under some other heading and/or it exists and is not widely disseminated because it is not compatible with current genetic theory. I would enjoy hearing from some of the experts on this subject.
It is fascinating trying to reconcile- 1)the assertions that mutate-select can easily explain changes in a variable like wing length, 2)the complete inability of proponents of current genetic theory to come up with even a marginally reasonable explanation/models of the phenomena, and 3)the great reluctance of these same people to even seriously consider seemingly obvious alternatives to current genetic theory. Fascinating.
Frances,
So you consider ‘scientific predictions’ to be soft evidence. That must leave you with a rather strange view of science.
Biological or intelligent design is far more complex than plain, simple, adaptive design. Most obviously, biological systems have the ability to evolve, to create novel solutions (adaptive solutions outside the range of current potential solutions), the ability to create new adaptive elements and systems, the ability to undergo dramatic transformations, and the ability to form complex structured solutions from simple elements. A biological design process, it appears, may have created life from non-life.
It is, IMO, useful to first expand our understanding of simple adaptive design before considering the complexities of biological design.
QUOTE: "Also the claim that (intelligent) design will be faster remains unsupported. In fact I could provide you with some examples in which RM&NS seems to have been much more productive in finding solutions than intelligent designers."
I claimed that a ‘directed design’ process could complete a step by step adaptive change much faster than an RM&NS. The mechanism I proposed could potentially go through the 10 step process in 10 generations using elementary bio-chemical mechanisms, relatively small populations and accelerated selection(selective breeding). An RM&NS model using realistic mutation assumptions, limited populations, and selective breeding would require at least 1000’s of generations to produce the same result. Furthermore, if the ‘redesign involved many different variables, the ‘directed design model could proceed at speeds very close to those achieved in a one variable model. The speed of an RM&NS based model would slow down dramatically as the number of variables increased. RM&NS, can not explain, model, or simulate anything but the simplest of one mutation changes.
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 12. August 2002 19:52
quote:
Warren posted:
You made a valiant, but obviously unsuccessful attempt to describe how a ‘simple’ 10 step change in size of a single variable might be accounted for using mutate-select genetic concepts.
You are too kind. In fact, it was a half-assed attempt jotted down in 15', based entirely on entry-level evolutionary theory - stuff you could find in any textbook (really!). I do agree though, that it already seems to beat in complexity and definition any mathematics you have presented in support of your own mathematical model - but that does not seem too hard, since we are still waiting for it.
quote:
A couple of key elements of your attempt should be pointed out. First, and most obviously, you failed completely to come anywhere close to the redesign speeds demonstrated by selective breeding.
Of course, that was an example based on the evolution of a quantitative trait under natural, not artificial selection. Artificial selection is an entirely different ballgame, since:
a) it relies heavily on pre-existing variation;
b) The fitness advantage of any favored trait is 100%.
Basically, MUCH faster.
Given however that your initial problem dealt with eagle wing length, I thought you were referring to evolution under natural conditions. Or maybe you are referring to some artificial eagle breeding experiments I am not aware of?
quote:
The central point of the alternative model I offered is to explain rapid rate of adaptive change that appears to occur in nature that and can be induced under artificial conditions.
You may want to discuss exactly what "rapid rate of adaptive change in nature" you are talking about. I think both Frances and I are trying to understand what exactly you have in mind. Is it eagle wings? How fast do you think eagle wings evolved? Is there any evidence from the fossil record, or are you just assuming it was really fast? Or are you talking about chihuahuas and other artificially derived breeds? In fact, most breeders use old-fashioned genetics to plan their breedings. The last geneticists who relied on "quickly adaptive" changes under selection for their breeding schemes were the Lysenkoists, in stalinist Soviet Union. Needless to say, it did not work.
quote:
In order to develop a mutate-select model to fit step by step changes you are forced to speculate on the existence of very complex genetic coding mechanisms. Mechanisms which would require huge numbers of genes and for which there is absolutely no evidence.
In fact, the model of genetic control of wing development I proposed was ridicolously simplified, compared to what we know of vertebrate limb development; there probably are several dozens genes that play important roles in making an eagle wing. I strongly encourage you to take a look at the references about developmental biology I listed in a post above. Unless, of course, you want to provide some molecular developmental evidence in favor of your alternative "single gene with 10 alleles" model.
quote:
As an aside, I find it ... somewhat difficult to believe that there isn’t a significant body of knowledge on the mechanisms underlying ‘small incremental changes in the variables controlled by developmental processes’. Such mechanisms would be relatively easy to analyze and of obvious direct interest to genetic theory. The information must either exist under some other heading and/or it exists and is not widely disseminated because it is not compatible with current genetic theory. I would enjoy hearing from some of the experts on this subject.
I imagine you are talking about genetic mechanisms here (for evolutionary models, see below). As I said before, there is an increasing body of genetic and molecular knowledge about quantitative trait loci (QTLs). However, it is only recently that the technology has developed to the point that we can genetically map and identify the genes involved in quantitative traits. Let me take a short digression and explain how this works.
Imagine a single-gene trait (your classic mendelian character). If you know the phenotype, and want to know the gene that underlies it, you have 2 strategies:
1) the candidate gene approach (that's a fancy term for "guessing"): that is, based on existing knowledge, you make educated guesses about what the gene or protein involved may look like (for instance, based on an already characterized biochemical pathway, or on another known mutant with similar phenotype), then go check the gene and see if it's mutated in carriers of the trait.
2) "positional cloning". This strategy is based on the fact that we can now track the transmission of sub-sections of chromosomes among the progeny of crosses using "molecular markers" that differ between individuals. That is, given two individuals with known phenotypes (e.g. one carrying the trait of interest, the other the "wild-type"), we can identify specific portions of their individual chromosomes using these markers, and see how often the trait in question and each one of the markers end up together in their grand-children, grand-grandchildren, etc. The more often the trait ends up associated with a certain marker, the more likely its chromosomal position is close to that marker (chromosomes get all mixed up at each new generation, so genes and markers far from each other tend to get separated). Once you statistically find out which marker your trait has segregated with most often, if your analysis reaches a certain statistical threshold you can be relatively sure that the gene for your trait maps in close proximity to that marker. Then, since you know where the marker maps, you can clone the piece of DNA surrounding it and go look for genes around there (say, one or 2 million base pairs, vs. the initial 3 billions you were looking at before mapping - a pretty good enrichment!), until you find the right one.
Now, I know this seems really straightforward ( ![[Wink]](wink.gif) ), and in fact it is (relatively speaking) for mendelian, single gene traits. However, if your trait is determined by multiple genes (i.e. QTLs), all contributing to it in small and different ways, the difficulties are increased enormously, both from the genetic standpoint (you need many more markers to reach significance, and much larger progeny cohorts), and mathematically/statistically, in terms of resolution of the various patterns of association. So, I can't really agree with you that QTLs must be "relatively easy to analyze" (they are actually the most difficult types of genes to look for - believe me, I am trying to find one right now), although I agree that they are very interesting in genetic terms.
Because of these difficulties, it has been only in the past 3-4 years that these studies have become feasible, and generally in easily breedable organisms: mostly drosophila, plants, mice. Again, because so many economically valuable phenotypic traits are quantitative (milk production, muscle mass and fat content, fertility etc), serious efforts are also being put in the quantitative genetics of some farm animals which otherwise would be a pain to study.
The references I gave you already are a good starting point, and provide a reasonable idea of the state of the subject; if you are indeed interested, give them a shot.
quote:
It is fascinating trying to reconcile- 1)the assertions that mutate-select can easily explain changes in a variable like wing length, 2)the complete inability of proponents of current genetic theory to come up with even a marginally reasonable explanation/models of the phenomena, and 3) the great reluctance of these same people to even seriously consider seemingly obvious alternatives to current genetic theory. Fascinating.
Fascinating indeed, especially because, while 1 is true (although I would say that RM/NS models are fully compatible with, not that they can easily explain, wing evolution), 2 certainly is not. For instance, here is a recent paper on the modeling of evolution of quantitative skeletal features in dogs(BTW, if you enter "quantitative trait evolution" into Pubmed, you'll find some 140 papers); here is a review on the quantitative genetics of morphometric variation. So, from the published record, it looks like biologists think current models are indeed doing a reasonable job, so far. Maybe, you can pick one of these papers, the dog skeleton one for instance, and show us its deep thoretical deficiencies. That should be relatively easy, shouldn't it. Any cogent argument would, I can assure you, win this biologist reluctance "to ... seriously consider seemingly obvious alternatives to current genetic theory". Promise.
[ 12 August 2002, 19:56: Message edited by: charlie d. ]
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 12. August 2002 23:28
Warren:
quote:
So you consider ‘scientific predictions’ to be soft evidence. That must leave you with a rather strange view of science.
Predictions hardly qualify as evidence Warren. Predictions are validated or refuted by evidence but they hardly make for evidence all by themselves.
QUOTE: "Also the claim that (intelligent) design will be faster remains unsupported. In fact I could provide you with some examples in which RM&NS seems to have been much more productive in finding solutions than intelligent designers."
quote:
I claimed that a ‘directed design’ process could complete a step by step adaptive change much faster than an RM&NS.
But it does not have to do it faster thus speed cannot be used as evidence of directed design processes.
quote:
RM&NS, can not explain, model, or simulate anything but the simplest of one mutation changes.
I am sure that you can back up this assertion thus I am looking forward to a better documentations of these claims.
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 13. August 2002 10:06
Charlie D:
This will probably sound naive, but I am wondering about a couple of technical points in your discussion of the genetics of eagle wing evolution.
If we posit 10 genes just for the wing, and assume 50,000 genes for the entire genome, that means only 5,000 independent "traits" (assuming the wing is typical in number of genes required). Doesn't this seem ridiculously small, considering the complexity at the cellular level alone, never mind the physiological level?
If we say that most or all genes are pleiotropic though (which is surely the case), doesn't that ruin your whole argument? Don't we then have a nightmarish complexity that makes such statistical reasoning impossible? In that case, don't we have to assume that different traits are somehow evolving in tandem (i.e., teleologically)?
Also, the figure 200,000 years for evolution of long wings---isn't that on the long side, given the well-documented "punctuated" character of the fossil record? Or even given the claim that one sometimes runs across that natural (not artificial) selection of comparable traits has actually been observed over extremely short time spans (the famous industrial melanism case, Hawaiian flies [if I remember correctly], and perhaps a few others)?
It seems to me that Darwinians want to use a lot of math to show what would happen under a certain set of unrealistically simplistic assumptions. There is no problem with the math; the problem is whether the assumptions throw any light on reality.
Also, what do you make of the recent efforts by many biologists (not philosophers, and certainly not creationists) to argue that genetic variability is under direct metabolic control, at least in some organisms? (I am thinking here of the NYAS anthology edited by Lynn H. Caporale, "Molecular Strategies in Biological Evolution.") And then there are the contributions from "developmental systems" (see the MIT Press anthology "Cycles of Contingency" edited by Susan Oyama et al.), behavior ("Baldwin effect," etc.), and cytoplasmic inheritance (Eva Jablonka and Marion J. Lamb, "Epigenetic Inheritance and Evolution").
In short, given all of these complexities and uncertainties and possible alternative explanations, don't you think that the simplistic assumptions underlying your mathematical model---which is unimpeachable GIVEN the assumptions---rob population biology of much of its explanatory value?
[ 13 August 2002, 10:19: Message edited by: James A. Barham ]
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 13. August 2002 11:26
James:
of course my model is ridicolously simplistic and utterly useless, if I may say so myself (in fact, I did, above). We simply do not have the knowledge to properly model wing evolution.
What we do know, however, is that evolution of complex quantitative traits (which longer wings most likely are) does not pose any specific theoretical obstacle to darwinian processes, and that natural selection is rather efficient at picking up and expanding adaptive variation within populations, whether pre-existing or newly generated by mutation. Both these conclusions are in stark contrast with warren's unsupported claim that traits like longer wings can simply not evolve by darwinian mechanisms (whether wings in fact did evolve that way, of course nobody can say right now).
As for your specific genetic/evolutionary comments:
- yes, many genes that would influence longer wings would likely also affect other characters (see the analysis of correlated dog skeleton variations in the paper I listed). The combined fitness of all the phenotypic changes is what NS will have to sort through (unless there is evidence that longer wing necessarily bring about some other counterselected change that would shift the adaptive balance, it's not a problem).
- Yes, evolution of longer wings would require coordinated evolution of other traits, but I don't see why that is necessarily teleological. As long as character 1 is not counterselected when it first appears alone, character 2 can appear sequentially. For instance, longer wings most likely require stronger pectoral muscles. However, an only slightly longer wing (10% of the final difference) may not have such a requirement, and still be useful (eg, for gliding); once slightly longer wings do become common, however, selective pressure will push for stronger pectorals, and so on. Thus, the 2 characters will end up evolving in tandem.
- No, 200 Ka is pretty much nothing in evolutionary terms, even for punctuated equilibria (which in any case are not necessarily the rule, especially for something so obviously gradual as evolution of longer wings). I have no clue what the eagle fossil record shows, though.
Again, the issue here is not: do we know with any degree of scientific certainty how longer wings evolved, based on darwinian mechanisms? The issue raised by warren is: do darwinian mechanisms grind to a halt when faced with something like evolution of longer wings? From all we know about developmental genetics and quantitative trait evolution, quite clearly they do not. Could other entirely hypothetical mechanisms also explain wing evolution? Sure, of course, but why should we prefer them over well-characterized ones?
I think, for the "problem" of longer wing evolution to become an issue at all, you either have to show that eagle length is under some very unusual genetic/developmental control, or that selection had to face some very unusual adaptive obstacle (for instance, that any even minor variation in wing length is counterselected), or that wing evolution somehow occurred with a historical pattern not consistent with darwinian mechanisms. At the current status of knowledge, however, evolution of longer wings in eagles is not a problem at all for evolutionary theory.
This is really it for me on this topic, though. warren can have the last word, if he likes.
[ 13 August 2002, 14:05: Message edited by: charlie d. ]
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 14. August 2002 09:43
Frances, Charlie, James
Quote Charlie: You may want to discuss exactly what "rapid rate of adaptive change in nature" you are talking about. I think both Frances and I are trying to understand what exactly you have in mind. Is it eagle wings?
I am using the wings of eagles redesign example to show that 1)there exist models/theories which are alternatives to and incompatible with Darwinian/neo-Darwininian models/theories, 2)there are rigorous scientific techniques for evaluating/comparing the Darwinian and non-Darwinian models and theories, and 3)there is ample rigorous scientific evidence clearly demonstrating that the non-Darwinian theories are superior.
Your inability to understand and accept these demonstrations is in part due to your expectations that such demonstrations must be very complex, must be based on new evidence that no one has seem before, and must rely on reference to a whole array of existing experts. Instead, my argument is based on plain, elementary scientific concepts, evidence that is readily available( but easily misinterpreted), and on avoiding so called experts. Let us for the moment be kind and assume that your inability to understand and accept the conclusions presented here is due to my failure to clearly express myself. Let me therefor make one more attempt at clarification.
A NON-DARWINIAN HYPOTHESIS
In my wings of eagles example, I presented a specific non-Darwinian model of evolutionary change that applied to one specific feature (evolutionary change in continuous change variables like wing length). Rather that a model relating to a specific set of evolutionary change processes, let me offer the following more general non-Darwinian design science hypothesis relating to evolutionary and adaptive change.
DIRECTED DESIGN HYPOTHESIS- Evolutionary and adaptive change processes in biological systems contain directed design elements, processes, components, and/or mechanisms which significantly increase the rate of evolutionary and adaptive change(for specific adaptive and evolutionary change problems).
Strictly speaking, the directed design hypothesis is a supplement to Darwinian theory. Evolutionary change, it could be argued involves both the Darwinian selection, diversity, and heritability and the directed design features implied by the above hypothesis. It has always been accepted that Darwinian theory is subject to modification and refinement. Whether the above hypothesis is a modification or a rejection of Darwin is, IMO, not terribly important.
On the surface, the directed design hypothesis would appear to directly contradict RM&NS and neo-Darwinian genetics. But even with respect to genetic theory, it is important to distinguish pop-culture rhetoric from ‘scientific’ concepts. If one takes a narrow, literal view of genetic theory, then the directed design hypothesis is a direct contradiction to RM&NS. If one takes the much broader view that RM&NS is a loose label covering all possible naturalistic explanations of evolutionary change, then the directed design hypothesis is simply a supplement to existing genetic theory.
The analysis offered here is in support of the directed design hypothesis. Whether the analysis provided is also viewed as ‘contradicting’ Darwinian or neo-Darwinian ‘theory’ depends on how narrow or broad your view of these theories. In reviewing this analysis, you must first answer whether you view the directed design hypothesis as compatible with Darwinian theory. What is your view?
A RIGOROUS SCIENTIFIC BASIS FOR EVALUATION
The best, most rigorous scientific basis for testing and validating a theory is 1)predict a quantifiable result, particularly an unexpected quantifiable result, and 2)make observations which support the predicted result. Using design science concepts, it is possible to produce quantifiable predictions and to demonstrate the validity of the predictions. Verifiable quantitative predictions are hard scientific evidence. If you reject this fundamental principle, then you have in effect rejected the scientific paradigm.
In evaluating the directed design hypothesis we are interested in evaluating for 1)a system with a measured minimum level of complexity, 2) the actual average or repeatable rate of change relative to 3)the random or trial and error rate of change. I am assuming here that you understand how to measure and/or approximate each of the these values. For the examples being discussed here, the calculations are fairly simple, and you should be able to generate acceptable approximations ‘in your head’.
In the discussion here we are considering ‘actual average rates of change’ of ‘less than 100 trials’ and ‘random trial and error speeds’ in excess of the 100’s of millions of trials. Directed design is not about a 2% increases in evolutionary speeds, but ‘million fold’ increases in speed( it is in fact easy to find examples where the increases are far greater than that).
The directed design hypothesis predicts that 1)there are instances of evolutionary change where the average repeatable speed will significantly or materially exceed the random or trial and error rate and 2)there will be(may be) identifiable processes and mechanism explaining the of accelerated rates of evolutionary/adaptive change. The speed predictions can be readily measured/tested by observations of biological systems. The existence or non-existence of hypothesized physical mechanisms can be tested by standard experimental techniques. These test processes are simple, basic, and rigorous. If you reject the proposed tests, then you reject (IMO) the basic principles of scientific analysis.
As I stated earlier, it is not logically necessary to view directed design as incompatible with Darwinian or neo-Darwinian theory. Quite obviously, directed design is incompatible with many of the pop culture interpretations of Darwinian theory.
THE OVERWHELMING DIRECT EVIDENCE SUPPORTING DIRECTED DESIGN
To begin, I would like to thank James for his discussion of some of the details of the research relating to this topic. The evidence supporting the existence of directed design in evolutionary processes is based on information that is well known and readily available to everyone with even a limited interest in the subject. My arguments and my analysis are based on interpreting or reinterpreting well known facts. It is, however, reassuring to know that there is support for at least some of my claims in some of the less well known literature.
Based on the criteria defined above, eagles wings provide ‘hard evidence’ supporting the directed design hypothesis. The claims/predictions relating to one of the potential mechanism underlying wing redesign, and the claim/prediction relating to survivability can not be currently answered, but there is plenty of available evidence to evaluate the issue of speed. Evolutionary redesign of eagles wings clearly occurs at a pace greatly in excess of the random mutate-select rate.
In order to evaluate the speed prediction, we need 1)a conservative(low end) estimate of minimum random mutate-select rate of wing evolution and 2)a conservative(high end) estimate of the maximum expected rate of change.
In order to obtain an estimate of the minimum mutate-select rate we need to consider variables like number of mutations, mutation rates, and selection rates. If, in doing such calculations you eliminate as a possibility the ‘one step miracle’ and require multi-stage gradual change, the minimum mutate-select speed increases dramatically as 1)the number of mutations increases, 2)the complexity of the mutation increases, and 3)as the force of selection decreases. Charlie’s estimate of 200,000 years per step for a 10 step process is optimistic. If you recognize that wing design involves many variables, and probably many more than 10 possible steps for each variable, then it becomes fairly obvious that ‘random mutate select processes could not possibly explain evolutionary redesign of wings. Using realistic assumptions, the ‘random mutate-select’ time required for wing redesign would quickly run into the billions of years.
For the discussion here, we can start with Charlie’s estimate of 2 million years for redesign from short wings to long wings via a random mutate-select process. It is, however, useful to also keep in mind that this is an extremely conservative estimate. More realistic estimates would be much longer.
The second part of evaluating evidence is determining a conservative (high end) estimate of the actual expected redesign speed. To answer this we can look to the fossil record. We can also look at how fast changes occur in the environments in which eagles live, the number of species currently mal-adapted or poorly adapted to the environments in which they live(the number of species currently under going gradual evolutionary change, and we can look to the rates of change which can be produced by artificial breeding. Given the available evidence, it is not unreasonable to estimate that rather dramatic redesign might occur in 100’s or 1000’s of years rather than millions or billions of years. It is, at the very least, obvious that that actually rates of evolutionary change are much faster than the random mutate select rate.
I am sure it is possible to do a much better job of documenting step by step the calculations and evidence relevant to directed design. It should, however, be obvious that the evidence exists. It should also be obvious that ‘directed design’ is not necessarily the sign some type of black magic. There are simple, logical, mechanistic explanations for many(if not necessarily all) forms of directed design.
As an interesting aside, it is, or should be, fairly obvious that ‘directed design’ plays an important role in human problem solving. A great deal of research has and is being done on how humans find solutions(or adaptive designs) to complex problems(from sets of possible solutions or designs. Elaborate search routines (often based on evolutionary computation concepts) have been developed in attempt to simulate the human problem solving techniques. As with ‘random mutate-select techniques’ the number of trials, or the amount of processing increases (speed of finding an adaptive design slows down) as the complexity of the problem increases. Actual human problem solving, as is easily observable, find solutions(adaptive designs) to many of these very complex problems very quickly. The speed and ease with which humans solve certain complex types of problems is and/or must be due to the use of directed design processes and mechanisms.
In conclusion, it is useful to note that ‘directed design’ while it provides a benefit to an organism, is not without its price. The ability to quickly find a directed solution is achieved by focusing or directing analysis to certain types or classes of solutions. This means that other types or classes of possible solutions are overlooked or ignored. As an interesting real world example, it could be argued that biologists are currently ‘directing’ their efforts to understand evolution change toward Darwinian and neo-Darwinian solutions. As a result of this ‘directed problem solving’ potential solutions in other directions are being ignored or overlooked.
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 14. August 2002 12:41
Warren
quote:
am using the wings of eagles redesign example to show that 1)there exist models/theories which are alternatives to and incompatible with Darwinian/neo-Darwininian models/theories, 2)there are rigorous scientific techniques for evaluating/comparing the Darwinian and non-Darwinian models and theories, and 3)there is ample rigorous scientific evidence clearly demonstrating that the non-Darwinian theories are superior.
I have shown that your model does not seem to be different from Darwinian models, that no rigorous techniques have been used to evaluate or compare your model and certainly your claim that non-Darwinian theories are superior would require at least an attempt to make such a demonstration.
So far I would say that you have not done a very good job at demonstrating this, in fact others have shown much better evidence to the contrary.
Let's focus on first demonstrating and then making the claims, it would really benefit our discussion if we were to follow such approach.
Perhaps as a side note you could explain to us what or who is doing this 'directed designing'?
[ 14 August 2002, 12:59: Message edited by: Frances ]
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 15. August 2002 10:04
The discussion here has come down to the validity of two statements:
1. Based on the concepts and principles of evolutionary biology, the evolutionary redesign of eagles wings is an example of evidence supporting Darwinian and neo-Darwinian theory. and
2. Based on the principle and concepts of design science, the evolutionary redesign of eagles wings provides evidence supporting the directed design hypothesis.
Given current knowledge, both statements APPEAR to be true, yet they contain a clear contradiction. By most generally accepted interpretations the first statement asserts that ‘the evolutionary redesign of eagles wings can be scientifically explained WITHOUT assuming the existence of directed design processes or mechanisms. The second statement asserts that ‘the evolutionary redesign of eagles wings REQUIRES the assumption of directed design processes and mechanisms’.
The very different conclusions produced by evolutionary biology and design science APPEAR to illustrate the existence of different sets of scientific standards and principles. The following are apparent or possible interpretations of the evidence relating to the redesign of eagles wings.
THE APPARENT EVOLUTIONARY BIOLOGY INTERPRETATION
Evolutionary biology asserts that there are (at least) two different (groups of) Darwinian/neo-Darwinian models or theories which can describe evolutionary redesign(the traditional gradual change Darwinian model and variations of the mutate-select neo-Darwinian genetic theory). The available data on redesign appears to fit one of the models(the basic Darwinian gradual change model). It does not currently appear to be possible to fit eagle redesign to any known form of ‘random mutate- select model’. Based on this ‘evidence’ and lack of evidence, evolutionary biology accepts the assertion or claim that ‘the evolution of eagles wings’ is compatible with current theory and (implicitly) it is unnecessary to assume the existence of directed design mechanism to explain wing redesign.
DESIGN SCIENCE INTERPRETATION
Based on 1)known techniques for measuring actual speeds of evolutionary change processes, and 2)known techniques for measuring the speeds of evolutionary change based on ‘random mutate-select’ processes in systems of known complexity, design science permits/requires testing of the hypothesis ‘the evolutionary redesign of eagles wings can be modeled, simulated and explained by a random mutate-select model". The evaluation of available evidence using these techniques falsifies the adequacy of the random mutate-select hypothesis. Design science permits/requires the formulation of the ‘directed design hypothesis’ as an alternative to the falsified random mutate-select hypothesis. The alternative hypothesis explains the test results which falsified the original hypothesis and produces additional testable predictions.
As described above, the design science interpretation of the evidence relating to the wings of eagles is, or appears to be based on the information currently available, logically consistent with the most rigorous scientific standards. The evolutionary biology interpretation, as outlined above, at the very least, ‘appears’ to be guilty of making and supporting assertions which are not justified by the facts. At the worst, the interpretation outlined above suggests the possibility evolutionary biology, or more accurately evolutionary biologists, are guilty of intentionally ignoring evidence which contradicts existing evidence.
To be fair, the evolutionary biology interpretation offered above is based on the views expressed here by proponents of Darwinian theory and may not accurately reflect the views of evolutionary biology. Furthermore, if and when the analysis offered here is reviewed by qualified biologists, it is possible even likely that their interpretations will be consistent with those generated by design science.
Some individuals may wish to question the logic, mathematics, and evidence underlying the ‘directed design hypothesis’. It should be noted that directed design as a hypothesis is essentially a trivial, obvious hypothesis. It would not be seriously considered except for the fact that RM&NS appears to suggest/predict it isn’t true. The power and prevalence of specific directed design mechanisms in evolutionary processes may come as a surprise to some people, but the existence of such mechanisms is obvious.
The obvious, almost trivial reality of evolutionary processes, is that they are very complex and involve a wide range of different processes and mechanisms. Many, probably most of these mechanisms serve to make evolutionary change more effective and more practical. Design science provides an array of techniques and tools for identifying, measuring and evaluating a wide range of different evolutionary and adaptive change processes and mechanisms. Evolutionary biology gives the impression of attempting to suppress analysis of processes and mechanism other than those identified by existing theories.
The analysis of the redesign of eagles wings provides an opportunity to illustrate the tools and techniques of design science. Probably far more important, the redesign of wings provides an opportunity to identify/evaluate the scientific standards and principles actually used in evolutionary biology and to identify/evaluate the scientific standards and principles which ‘appear’ to used in evolutionary biology.
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 15. August 2002 12:23
Warren,
You are still making claims about speed and inadequacy of Darwinian mechanisms but seem to have failed to address the most relevant aspect, supporting these claims with historical and biological data for the eagle wing. Your 'logical' analysis is thus based on faulty premises and incompleteness thus your conclusions are immediately suspect. Thus I asked you for an effort on your part to support your claims but you continue down a path that has little to offer to the discussion. Your claims about RM&NS for instance still remain unsupported. That by itself fully undermines your analysis.
So let me repeat my question:
You have made a lot of claims about RM&NS, evolution as well as the speed of evolution wrt the eagle wings but I have not seen any hard evidence/data that would seem to support your arguments. In fact the only quantitative attempts have been from your opponent(s) who have shown that your arguments seem to be suffering from a simple though devastating problem, lack of foundation in reality.
I do realize that arguing based on logical premises is far easier than arguing based on the facts but such arguments seem to be quite limited in their relevance to reality.
I am looking forward to your future postings in the hope that you will provide for some relevant data.
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 15. August 2002 15:10
Francis,
You request for ‘hard evidence’ is nothing more than a rhetorical effort to avoid discussing the issues involved. Even moderately realistic estimates of the ‘random mutate select’ time requirements would be in the millions or billions of years. Realistic estimates of ‘maximum rates would be less than 100,00 years, probably in the 1000’s or 100’s of years. Unless you have a supportable reason to doubt either of these estimates you are raising issues that have absolutely no material impact on the subject being discussed.
If as you suggest, my analysis is based on faulty premises then point out these faults. Charlie spent a great deal of time attempting unsuccessfully to demonstrate that a ‘random mutate select’ model could generate realistic rates of evolutionary change. His effort was a waste of time. It is ‘obvious’, if you know any math at all, that such efforts are bound to fail. Wing designs are very complex and very precise. Mutate-select models with very slow realistic mutation rates and low realistic natural selection rates grind to a halt very, very quickly when attempting to address complex problems. That’s obvious.
It is equally obvious from discussions of selective breeding, from the fossil record, from observation of the very limited number of species currently undergoing ‘gradual adaptive change’, and from observations of the rate of climate changes, it is ‘obvious’ that changes such as wing redesign must occur very rapidly. Lots of people have put together lots of evidence relating to this topic. It would be a complete waste of time for me to rehash and reevaluate an existing body of knowledge.
If you have reason to actually doubt the obvious estimates being used here then actually present them. Suggesting that an argument looses credibility because it does not endlessly rehash the obvious is, IMO, ridiculous.
Quote: "In fact the only quantitative attempts have been from your opponent(s) who have shown that your arguments seem to be suffering from a simple though devastating problem, lack of foundation in reality."
Charlie, the only opponent who has attempted to identify my ‘lack of foundation in reality’ failed miserably and ended up demonstrating the very assertion he was attempting to disagree with. If, as is always possible, I missed either evidence or a valid argument demonstrating ‘a lack of foundation in reality’ then present it. If you can’t support your claim then I must assume it is simply rhetoric.
Quote: I do realize that arguing based on logical premises is far easier than arguing based on the facts but such arguments seem to be quite limited in their relevance to reality.
I would conclude exactly the opposite. I see lots of examples of people trying to use long lists of facts as a substitute for logic and reality. I see very few people being capable of tying those facts together logically to identify reality.
The issues in evolutionary science is not facts, we have lots of facts, but interpreting the facts to form some type of reasonable reality. The ‘directed design hypothesis’ is a simple observation that 1)the facts relating to the complexity of evolutionary problems, and 2)the facts relating to the speed with which living organisms solve those problems, is not compatible with 3)a theory that says the problem solving must follow a random process. I think such an observation bears a great deal of relevance to reality.
The reality which you, Charlie and neo-Darwin theory seem to be arguing is 1)the theory says that problem solving must be random therefor 2)the facts must show that the speed of evolutionary change must be consistent with the rate expected from a random process. Quite frankly, it is the neo-Darwinian arguments that, IMO, ‘seem to be quite limited in their relevance or reality’.
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 15. August 2002 17:12
Charlie said he was done with this thread, but maybe I can entice him back for one more round.
Charlie wrote:
"Yes, evolution of longer wings would require coordinated evolution of other traits, but I don't see why that is necessarily teleological. As long as character 1 is not counterselected when it first appears alone, character 2 can appear sequentially."
When I say that such coordination is "teleological," I just mean that it displays the general power of the organism to harness the events within it for its own good, its own self-preservation. I just mean that we have to posit a holistic or top-down form of organization, that a reductionist or exclusively bottom-up view of the organism does not jibe with the facts.
Then, if this internal or immanent teleology or self-organizaing power or whatever you want to call it on the part of the organism be admitted, I am taking one more step and claiming that much of the seeming explanatory power of the theory of natural seleciton in fact trades on the implicit, unrecognized explanatory power of this holistic self-organizing power.
Without the assumption that traits evolve somehow in tandem, Warren's and Dembski's and many other people's criticism of the RM/NS mechanism are on target. It is simply unthinkable on probabilistic grounds (or statistical-mechanical grounds, if you prefer). But with it---that is, with the assumption of an internal self-organizing power---then we no longer have a reductionistic RM/NS mechanism. Rather, we have a nonlinear dynamical system with its own inherent propensities and powers. And that is now the real explanation for evolution, not RM/NS.
In short (and this recapitulates many discussions I've had with Frances on other threads), Warren's and my complaint is not that evolution is impossible. It is that evolution is impossible without presupposing an internal teleological principle of coherence and coordination constantly at work.
[ 15 August 2002, 17:18: Message edited by: James A. Barham ]
IP: Logged
|
|
|
Printer-friendly view of this topic