|
Author
|
Topic: Non congruent phylogenies
|
yersinia
Member
Member # 324
|
posted 24. December 2002 18:28
Random brief points on Christmas Eve:
1) I'm quite sure, based on what I know about common descent and cytochrome C, that the differences in the aa sequences of cytochrome C between bats, and between bats and other placentals, are trivial. It is just a few substitutions at most, and I'd wager that what few changes exist are small changes (hydrophobic aa replacing hydrophobic aa).
2) Hemoglobin on the other hand is another story. I'm not even sure if frogs have the four-subunit hemoglobin that mammals have. Certainly oxygen uptake processes differ rather alot between the two.
I guess I'm still confused about what is being proposed in this thread. Hunter has argued in various places in the thread:
a) That the oft-cited congruities between independently-derived trees are not so impressive after all
b) That ID predicts tree congruence just like common descent
c) That assessing the probability of tree congruence occuring by random chance (or "random design", whatever that it) is inappropriate, even though he has put forward no alternative null hypothesis.
d) That all of this evidence indicating that the sequence-to-function mapping is a (very) many to few relation can be brushed aside until we have an exhaustive understanding of the sequence-function relationship.
e) That "design transplants" (incongruence) are, just like congruence, also predict by ID. This is supported with some muddled statements about marsupial-placental convergences, without a shred of evidence that *any* genetic information has been transplanted between the two groups, and lots of evidence that the similar adaptations are independent solutions to the same ecological problems, resulting in superficial similarity in gross structure but nothing that would make a mammalian paleontologist confuse placental and marsupial skulls (and also nothing that would independently drive the DNA sequences towards convergence).
...and several other things which are even more confusing. Suffice it to say that this does not support his contention that the biologists are the ones who haven't "thought this out."
If the kinds of lateral information transfer that Hunter appears to think occurred in the history of multicellular animals had actually occurred, then we would see would see wildly different trees from different data sources, like we do (in some cases) with prokaryotes. But we don't see this.
I, however, have Christmas duties to attend to, and will withdraw for the next few days. Merry Christmas!
IP: Logged
|
|
Cornelius G. Hunter
Member
Member # 81
|
posted 24. December 2002 18:45
Argon, in your recent post in reply to John you wrote:
"A design framework that doesn't incorporate common descent has a difficult time explaining the relationships. Remember, morphologies, protein structure, protein sequences, DNA sequences and *time since separation/'creation'* all tend to produce overlapping trees. The 'time since separation' aspect is particularly difficult to square without descent + modification as a chief mode of design. The overlapping trees are also difficult to harmonize without common descent."
This is not clear to me, and I would appreciate it if you could provide your reasoning for the above conclusions, especially your first sentence. Don't feel like you have to provide the backup evidence (though that would be fine too). If you could just lay out the steps of your reasoning that would help. The reasoning is more important than the supporting evidence. For example, why are "overlapping trees" difficult to square without descent + modification?
Also, if you could make your terminology clear that would be appreciated (eg, what are overlapping trees? I assume you mean trees with similar topology).
Thanks so much.
--Cornelius
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 25. December 2002 11:19
Measuring incongruence
Responding to Yersinia's remarks about the relative size of incongruences and similarities of tree structures induced from data and that congruence and incongruence between induced trees are not "either/or" relationships, Hunter writes quote:
True, but they sometimes are. You write: "the interrelationships of different groups of bats is a pretty trivial issue in the context of vertebrates or animalia. If the microbats grouped most closely with anthropoid apes, and the macrobats with giraffes, then we'd have a significant disagreement." This tells me that evolution is quite flexible and can sustain a tremendous quantity of phylogenetic anomaly. As I suggested, this is what I suspected. While you say this bat example is pretty trivial, it has the consequence of causing echolocation to have evolved twice. So here we have an example which not only can we apply the usual statistical measure of significance and confidence to, but we can also see a significant evolutionary consequence.
The bioinformatics and biostatistics literature has not ignored this issue. See, for example, this paper and this paper and especially this paper for examples of that interest.
Congruence/incongruence of tree structures, if not a ratio-scaled variable, is at least a graded or interval variable such that it makes sense to measure greater or lesser congruence and therefore to use congruence as a dependent variable in testing hypotheses about the generator(s) of the trees under analysis. Hunter seems to agree that one can measure and evaluate relationships among tree structures when he writes of Yersinia's term "overlapping trees" quote:
I assume you mean trees with similar topology.
That's how I interpreted Yersinia's phrase, and Hunter's use of "similar" presupposes some sort of relative similarity measure or judgement.
The question at issue in the OP and subsequent posts is not whether biological groups form nested hierarchies that can be represented as tree structures. That is uncontroversial. The questions have to do with various ways of measuring the groupings of species (in the present instance) that form the nested hiearchies - morphological versus molecular in the present instance - and what implications they have for hypotheses about the generating process(es) and for inferred relationships of common descent. Do different data produce (radically or trivially) different trees, and what do the differences mean for theories of the generators of the structures?
To assert that the morphological and molecular trees do not match precisely is to pose questions for research given that the mismatches occur within the context of larger similarities of tree structures. Yersinia remarks that quote:
A design framework that doesn't incorporate common descent has a difficult time explaining the relationships. Remember, morphologies, protein structure, protein sequences, DNA sequences and *time since separation/'creation'* all tend to produce overlapping trees. The 'time since separation' aspect is particularly difficult to square without descent + modification as a chief mode of design. The overlapping trees are also difficult to harmonize without common descent.
I'd word that slightly differently, saying something like accounting for these similarities in an ID context without invoking common descent as evolutionary biology understands it requires considerable development of hypotheses about the intentions, methods, and proclivities of the hypothesized designing agency(ies). Those hypotheses require a design theory that is considerably more clearly articulated than anything I have ever read in the ID literature.
Explaining incongruences
In the OP Hunter wrote quote:
To be frank, I suspect most (all?) evolutionists have not thought this out, and there is no accepted precise theoretical formulation of Step 1, and that they are "winging it," for example when new non congruent phylogenies are discovered. I have seen several explanatory mechanisms used in such cases, they include:
a] the molecular data may have sufficiently diverged that the comparisons between them are not valid,
b] there may be too much noise in the molecular data,
c] lateral gene transfer has altered the molecular data,
d] the molecular data may be biased by the particular set of species under study.
e] Estimation errors can arise from data sets with many species
f] molecular data ought not to be equally weighted, [eg, segments influencing protein function should be assumed to be more important than the other segments, or perhaps segments influencing the protein structure should be more heavily weighted, or perhaps hydrophobic segments should be deweighted, etc]
My point is not to say explanatory mechanisms are out of bounds or that complicating factors should not be expected, but merely to raise the question: At what point does the use of these explanatory mechanisms become ad hoc and do we consider the Step 1 in the syllogism falsified? This evidence is prominent in the evolution literature, yet the details of just what is being claimed do not seem to be available.
At least one of the papers Hunter cited as displaying incongruent trees spends some considerable space on examining (not merely invoking as ad hoc handwaving) several explanations for the apparent incongruence. "Winging it" is an unjustifiably pejorative characterization of their examination of the question.
Hunter's remark that the Teeling, et. al., data seem to imply quote:
But now the new molecular phylogenies call for a different arrangement. If they are correct then echolocation must have evolved more than once, independently, in different bat species.
is worded more strongly than I think justified. In particular, the word "independently" must be understood in the context of a similar sensory adaptation arising twice starting from very similar lineages. It is worth noting in this context that marine mammals have also evolved echolocation capabilities, and that object location capabilities based on electromagnetic radiation (e.g., light) have independently evolved dozens of times in phylogenetic lineages much less similar than those involved in the (still hypothesized on the basis of one data set) two times for echolocation in bats. I can think of some ways of testing the "independent" evolution of echolocation in the two groups of bats that I don't see in the referenced papers and that I suspect would shed light on the observed difference between the morphological and molecular trees. As and if I have time I'll look around to see if some of those studies have already been done.
Learning from data
Finally, Hunter wrote quote:
My point was that Darwinists have been at this for many years and have crafted the best arguments possible. I appreciate the fact that there are other interpretations of common descent (emphasis added).
Not at all. Evolutionary biologists have crafted the best arguments possible given the data and knowledge currently available; as and when that inevitably changes, as one learns more (for example about the phylogenetic relationships among various bats) one can craft better explanations. 40 years ago the molecular trees could not have been constructed, say nothing of inform theory and test hypotheses about evolutionary processes. That is the notion of "progress" in science as I understand it. "The best possible at the moment," my old philosophy of science professor used to say.
RBH
IP: Logged
|
|
Cornelius G. Hunter
Member
Member # 81
|
posted 25. December 2002 15:56
RBH:
A couple of points of clarification. First, I believe you are confusing Argon and Yersenia in some of your responses, though your points still stand.
Second, Yersenia pointed out congruence vs incongruence is not an either/or question. This, of course, goes without saying. I had an editing (cutting/pasting) error in my response and my first sentence (ie, "True, but they sometimes are") was erroneously not deleted. Please strike that sentence.
Now, on to the specifics of your post. I am perplexed about the relevance of the three papers you cited. Yersenia had stated that "the interrelationships of different groups of bats is a pretty trivial issue in the context of vertebrates or animalia " and I had responded that there was more to the story. My point was that:
1) Yes, the difference can be made to appear trivial by placing it in the larger context or by comparing it with the random-design null hypothesis. A case of levitation could be made to appear trivial by comparing to all the examples of falling objects.
2) But there are biologically-relevant comparisons available to us and in fact, in the bat example, the point at issue was not so much the conflicting characters but rather the biological implications of the new best-fit phylogeny. Specifically, the new phylogeny called for the conclusion that echolocation must have evolved more than once.
3) I also pointed out that, in general, biologically-relevant metrics (such as the number of lateral gene transfers required, number of convergent mutations/adaptations required, etc.) provide meaningful evaluation criteria.
I fail to see how the three papers you cite are relevant to any of this. The first paper documents a database storage and retrieval project for storing phylogenetic trees from the literature. As part of the work they address the problem of measuring the distance between different trees.
The second paper is an evaluation and comparison of three different methods for analyzing and structuring gene expression data; namely, hierarchical clustering, k-means clustering, and self-organizing maps.
The third paper presents a new method (which I am not familiar with) for matching and comparing hierarchical structures which presumably could be used for comparing phylogenetic trees.
You then write: "Congruence/incongruence of tree structures, if not a ratio-scaled variable, is at least a graded or interval variable such that it makes sense to measure greater or lesser congruence and therefore to use congruence as a dependent variable in testing hypotheses about the generator(s) of the trees under analysis. Hunter seems to agree that one can measure and evaluate relationships among tree structures …"
Yes, I certainly agree. You then wrote:
"The question at issue in the OP and subsequent posts is not whether biological groups form nested hierarchies that can be represented as tree structures. That is uncontroversial. The questions have to do with various ways of measuring the groupings of species (in the present instance) that form the nested hierarchies - morphological versus molecular in the present instance - and what implications they have for hypotheses about the generating process(es) and for inferred relationships of common descent. Do different data produce (radically or trivially) different trees, and what do the differences mean for theories of the generators of the structures?"
Well said. You later wrote:
"At least one of the papers Hunter cited as displaying incongruent trees spends some considerable space on examining (not merely invoking as ad hoc handwaving) several explanations for the apparent incongruence. "Winging it" is an unjustifiably pejorative characterization of their examination of the question."
I'm sorry my "winging it" comment appeared to be an unjustified pejorative. And I certainly grant that evolutionists at times do seriously address the question of how to explain anomalies. I do, however, often see explanatory devices used quite freely. In fact, I think this is legitimate given that it occurs within the tradition of normal science and in light of the fact that the theory of evolution does not make highly specific claims but allows for great flexibility in sub theories. It seems that never (in the tradition of normal science) is evolution presented as something that may not be true. No matter what anomaly is being reported on, one always see the explanatory device given (eg, in the Raymond study it was massive LGT on a level heretofore not considered).
In the 1960s Karl Popper stated that he felt evolution was not falsifiable. This created a stir and evolutionists sought to deny the claim. David Penny gave this some thought, and presented phylogenetic analysis as a means of falsifying evolution [David Penny, L. R. Foulds, M. D. Hendy, “Testing the theory of evolution by comparing phylogenetic trees constructed from five different protein sequences,” Nature, 297 (1982) 297]. Penny said that high congruence was evidence for evolution, and that lack of congruence would serve to falsify evolution. Penny did not specify any precise statistical criterion for falsification. Now, 20 years later, we have plenty of data, ranging from high congruence to very weak congruence, depending on which species/character set and method one chooses.
Of course, there is no such thing as congruent trees and non-congruent trees as it all is a relative measure. So one defense is just to deny that a sufficient level of non congruency has been found. Another is to use random-design as the null hypothesis. Both of these give evolution a more than generous amount of wiggle room, and I think falsifies Penny's claim. That said, I'd prefer to avoid in this thread the issue of what constitutes falsifiability, and related philosophy of science issues. Let's keep the discussion manageable; namely, within the bounds of how phylogenetic analysis bears on evolution and ID. I think all of us can pretty much agree on the nature of the empirical and analytical results that we have in hand. What I think bears exploring is how we ought to interpret those results.
Next, you write:
***************************
--------------------------------------------------------------------------------
Hunter: But now the new (bat) molecular phylogenies call for a different arrangement. If they are correct then echolocation must have evolved more than once, independently, in different bat species.
--------------------------------------------------------------------------------
is worded more strongly than I think justified. In particular, the word "independently" must be understood in the context of a similar sensory adaptation arising twice starting from very similar lineages. It is worth noting in this context that marine mammals have also evolved echolocation capabilities, and that object location capabilities based on electromagnetic radiation (e.g., light) have independently evolved dozens of times in phylogenetic lineages much less similar than those involved in the (still hypothesized on the basis of one data set) two times for echolocation in bats. I can think of some ways of testing the "independent" evolution of echolocation in the two groups of bats that I don't see in the referenced papers and that I suspect would shed light on the observed difference between the morphological and molecular trees. As and if I have time I'll look around to see if some of those studies have already been done."
*****************************
Yes indeed, I also use electro-magnetic field sensing in fish as an example of high complexity in biology. No doubt, as you say "object location capabilities" must have evolved multiple times. I do not, however, see this as mollifying the point at issue but rather exacerbating it. Surely, it would be \it{petitio principii} to use the existence of echolocation in other species as justification for the multiple evolution of the bat echolocation. At best, I believe, you can argue from an appeal to authority (ie, echolocation evolving twice is OK because evolutionists have already accepted the multiple evolution of equivalent levels of complexity). But even this misses the point that the echolocation systems we find in bats is highly similar. Evolutionists have less difficulty in accepting the evolution of completely different systems than the repeat evolution of very similar systems because they see natural history of a process where contingency dominates necessity. In other words, the challenge of complexity needs to be distinguished from that of duplicated complexity.
I am, of course, not saying evolutionists cannot explain duplicate complexity. It is simple for them to explain it as the evolutionary process responding to a similar niche in a similar way. But at this point it becomes obvious there is very little it cannot explain.
Finally, you write: "… accounting for these similarities in an ID context without invoking common descent as evolutionary biology understands it requires considerable development of hypotheses about the intentions, methods, and proclivities of the hypothesized designing agency(ies). Those hypotheses require a design theory that is considerably more clearly articulated than anything I have ever read in the ID literature."
I agree that more articulation of ID will be helpful; however, it seems to me that your implying an additional burden on ID (as opposed to evolution) that doesn't seem warranted. As I have discussed there are plenty of anomalies from the evolutionary perspective and evolution has tremendous wiggle room given their use of the random-design null hypothesis.
--Cornelius
IP: Logged
|
|
Cornelius G. Hunter
Member
Member # 81
|
posted 25. December 2002 18:21
I think I figured out how to do quotes.
quote:
A design framework that doesn't incorporate common descent has a difficult time explaining the relationships. Remember, morphologies, protein structure, protein sequences, DNA sequences and *time since separation/'creation'* all tend to produce overlapping trees. The 'time since separation' aspect is particularly difficult to square without descent + modification as a chief mode of design. The overlapping trees are also difficult to harmonize without common descent. -- Argon
While waiting for Argon to explain his reasoning behind this claim, I thought I'd lay out my ideas to help stimulate thought and discussion. I'll refer to Argon's claim as "Claim A." It seems to me that the reasoning for Claim A must take the form of "the designer could not do it this way." Such a premise, it seems to me, is impossible to prove, so the argument must take a probabilistic form, ie, it is highly unlikely that the designer would do it this way. I call this the "cosmic conspiracy" form of argumentation, ie, yes the designer could have done it this way, but, because it is so unlikely, it would be tantamount to a conspiracy because it would be done in such a way as to make it appear that there was no design involved.
How can phylogenetic analysis, and in particular the high congruence observed in many character/species sets, support Claim A? Obviously designed objects will contain many correlated characters, ie, they will produce high congruence. So to support Claim A we must argue there are other characters that cannot be correlated, or at least are highly unlikely to be correlated. But we find uncorrelated characters in biology, so the argument must say either:
i) there are a great deal more such characters that ought not be correlated if designed than we find in biology, or
ii) there are certain characters that ought not be correlated if designed which we find to be correlated.
I believe (i) will be difficult, so let's look at (ii). In this thread, and elsewhere we see protein sequences, especially in large protein families, used as an example of design which ought not be correlated if designed. I'm running into two problems with this claim.
1) Given our rather incomplete knowledge of how the cell works it is difficult to support the claim that no functional purpose is served by the sequence variations. From a design perspective, it seems awfully coincidental that these sequences just happen to be highly congruent with the other characters which would be expected to form congruent phylogenies. We should also keep in mind the notoriously faulty record of the past claims of lack of reason or function for a design. As I stated earlier, I would be surprised if, as we learn more about whole cell operations, we do not find reasons for the sequence variations we observe. In fact, in my hypothetical frog multiple gene transplant experiment, even evolutionists consistently tell me they would expect the frog not to fare well. Clearly they believe the sequences are not completely arbitrary. Also, I note that though evolutionists have always claimed lack of function as evidence for their theory, the subsequent finding of function has never been considered as a falsification. For example, this quote is typical:
quote:
When structures undergo a reduction in size together with a loss of their typical function, that is, when they become vestigial, they are commonly considered to be degenerate and functionless. Simpson has pointed out that this need not be true at all: the loss of the original function may be accompanied by specialization for a new function. Thus the wing of penguins has become reduced to a point that will not permit flight, but at the same time it has become a highly efficient paddle for swimming. The wings of rheas, ostriches, and other running birds are also much reduced, and have been described as “at the most still used for display of the decorative wing feathers.” But Simpson has observed that the rheas, when running, spread the wings and use them as balancers, especially when turning rapidly. It seems quite probable that this is true of the running birds generally. [Edward O. Dodson and Peter Dodson, Evolution: Process and Product, p. 51, D. Van Nostrand Company, 1976.]
2) Let us say that in 20 years, after our knowledge of cellular operations has increased tremendously, we conclude that protein sequences are indeed arbitrary. I am still at a loss to understand why God ought not use congruent designs for the sequences. At best, it seems to me, the ID critic can point out that ID has no reason to prefer these sequences over any others (from within the set of viable sequences). That would certainly be valid but comes nowhere close to supporting Claim A.
Any thoughts?
--Cornelius
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 25. December 2002 20:31
Hunter wrote quote:
I fail to see how the three papers you cite are relevant to any of this. The first paper documents a database storage and retrieval project for storing phylogenetic trees from the literature. As part of the work they address the problem of measuring the distance between different trees.
The second paper is an evaluation and comparison of three different methods for analyzing and structuring gene expression data; namely, hierarchical clustering, k-means clustering, and self-organizing maps.
The third paper presents a new method (which I am not familiar with) for matching and comparing hierarchical structures which presumably could be used for comparing phylogenetic trees.
The three papers I mentioned (from a very large set of possible references) were intended only to make the point that the question of tree structure congruence has a large and varied literature, and is a useful area of knowledge for questions of phylogenetic relattionships inferred from molecular or morphological data. It was in large part stimulated by the sentence that was struck, and I accept that.
Hunter then wrote quote:
I'm sorry my "winging it" comment appeared to be an unjustified pejorative. And I certainly grant that evolutionists at times do seriously address the question of how to explain anomalies. I do, however, often see explanatory devices used quite freely. In fact, I think this is legitimate given that it occurs within the tradition of normal science and in light of the fact that the theory of evolution does not make highly specific claims but allows for great flexibility in sub theories. It seems that never (in the tradition of normal science) is evolution presented as something that may not be true. No matter what anomaly is being reported on, one always see the explanatory device given (eg, in the Raymond study it was massive LGT on a level heretofore not considered).
Perhaps I'm misunderstanding. If I understand the structure of evolutionary theory correctly, it does in fact have a sort of umbrella set of principles, on the order/level of common descent, mutation as a root source of variation, and a variety of means of dispersing novelty (sexual mating, recombination, LGT, etc.) through a population, and natural selection which filters the varieties of novelty by sculpting the variation offered up by mutations of various sorts, allowing or inhibiting their dispersal in the population through time.
Then there are what Hunter calls "sub-theories," which focus on more specific circumstances, offering testable hypotheses about smaller and more restricted classes of events. For example, the "sub-theory" concerned with endosymbiosis offers an account of the occurrence of certain sorts of intra-cellular structures and processes that are otherwise difficult to understand. It's analogous to having a general theory of gravitation, and then invoking sub-theories invoking air resistance to explain why feathers fall more slowly than lead shot in a 1-g field on earth. Then one explains (and tests) hypotheses about "air resistance" in terms of variables like density of the medium (all else equal, feathers fall faster in less dense air) and surface area/mass relations of physical objects in general. I doubt very much that a physicist who observed an object falling at half the expected rate in a 1-g field would subsume it under "noise"! Developing a theoretical structure in the "normal science" mode (though one can argue with Kuhn on this) seems to be largely the development, testing, and elaboration of such "sub-theories."
I agree that philosophy of science questions could take this thread far afield, though in my earlier posting I was reacting (without specifically mentioning it) in part to the 'falsification' remarks in the OP. A while back, I don't now recall where, I made the suggestion that anyone who uses the term "falsification" in a posting about science should be required to sign an affidavit attesting that the writer has (at least!) read both The Logic of Scientific Discovery and Conjectures and Refutations, as well as Lakatos's The Methodology of Scientific Research Programmes. (IMLTHO, Kuhn can be left FAR aside!) ![[Smile]](smile.gif)
Hunter writes that quote:
Yes indeed, I also use electro-magnetic field sensing in fish as an example of high complexity in biology. No doubt, as you say "object location capabilities" must have evolved multiple times. I do not, however, see this as mollifying the point at issue but rather exacerbating it. Surely, it would be \it{petitio principii} to use the existence of echolocation in other species as justification for the multiple evolution of the bat echolocation. At best, I believe, you can argue from an appeal to authority (ie, echolocation evolving twice is OK because evolutionists have already accepted the multiple evolution of equivalent levels of complexity). But even this misses the point that the echolocation systems we find in bats is highly similar. Evolutionists have less difficulty in accepting the evolution of completely different systems than the repeat evolution of very similar systems because they see natural history of a process where contingency dominates necessity. In other words, the challenge of complexity needs to be distinguished from that of duplicated complexity.
My intent in mentioning the other lineages in which various sensing systems (electromagnetic radiation includes light, btw!) have evolved was not to appeal to authority; it was to make something like the same point Hunter is making, namely that sensory systems come in a variety of kinds, but also the (hypothesized) fact that two very similar sorts of bat echolocation systems may have evolved in what are also very closely related (on other grounds) lineages does not seem to me to justify the adverb "independently." As I mentioned, other tests of whether the two echolocation systems are separate innovations in the two bat lineages identified in the one data set are possible, though I haven't yet had time to look for appropriate research. The repeated occurrence of a similar trait in two very closely related lineages is less surprising than if it occurred in two very distantly related lineages. The latter seems to me to be more in need of explanation than the former (and in aid of that explanation we have the hypothesis of common descent and the successful tests of that hypothesis in a number of subdisciplines). Hence my remark about "independently" not being really applicable to the bat echolocation data. "Independent" is an over-statement given the very similar (bat) substrates giving rise to the hypothesized two occurrences.
Remarks on Null Hypothesis Testing
This question actually first piqued my interest in this thread. Posing and rejecting a random null using a uniform PDF, of course, addresses only the simplest possible question: Is the phenomenon in question random? In the present context that seems to me to be a trivial question with an unexceptionable answer: Nope! That does not seem to me to be very interesting and is uninformative about mechanisms that might give rise to the phenomenon.
The more interesting questions arise when one poses alternative substantive hypotheses and attempts to distinguish among them. In general, my view is that 'orthodox' statistical hypothesis testing is not a terribly useful approach to characterizing and interpreting differences in data and theories. Rejecting a random null is less interesting (in science, at any rate) than exploration, and the statistical hypothesis-testing framework is not well adapted to exploring and extending knowledge. I see Bayesian approaches and likelihood estimation and effect size estimation as more useful methods for the difficult business of doing real scientific research. Only very rarely is there an opportunity for an experimentum crucis, a critical experiment that allows us to unambiguously distinguish between the predictions of two clearly framed hypotheses. Two examples that come immediately to mind are Eddington's observation of the (apparent) displacement of a star's measured position during a solar eclipse that discriminated between Einsteinian and Newtonian gravitational theories, and Michelson and Morley's test of the hypothesis of the existence of the ether as a medium for the propagation of electromagnetic radiation.
Much more common is the messy and difficult business of exploring a domain of phenomena guided by hypotheses or maybe just conjectures that range from ill-formed and vague to well-formed and definite; attempting to characterize and measure and manipulate the relevant variables and tease out their interactions; dealing with noisy data (where sometimes the noise is the data); and trying to build out of the various and sometimes contradictory findings a coherent account. A satisfactory account will first explain the data, where "explain" means (at least) that the account provides criteria for class membership assignment for a phenomenon ("This instance is a member of that class of instances") and identifies the relationships among classes that lead to the phenomenon's observed behavior. Scientific theories explain relationships among classes, not individual instances. Individual instances are explained by first identifying the class membership(s) of the instance, and then referring to laws and principles applicable to the class. The account will also provide criteria that allow us to classify and explain new instances of phenomena, instances not heretofore observed. Finally, the account will provide guidance concerning where and how to look for new phenomena and classes: it will generate a research program. That's how I see it now, at any rate. I'm currently (and concurrently) reading Mayer's What Evolution Is and Crump's A Brief History of Science in an effort to clarify some of the things I've held as axioms for several decades.
I'll respond to the most recent Hunter posting as (and if) I have something coherent to say.
RBH
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 25. December 2002 21:43
I think the problem so far with extending the issue to intelligent design is that we lack a coherent model for how intelligent designers would go about 'designing'. Mr Hunter suggests that
quote:
Obviously designed objects will contain many correlated characters, ie, they will produce high congruence.
versus
quote:
Let us say that in 20 years, after our knowledge of cellular operations has increased tremendously, we conclude that protein sequences are indeed arbitrary. I am still at a loss to understand why God ought not use congruent designs for the sequences.
Can we truely 'know' enough about intelligent designers and their motives to propose a non ad hoc hypothesis of intelligent design in biology? So far the approaches seem to be somewhat ad hoc and arbitrary in that it seems to suggest that intelligent design can in principle explain anything since we have no ways to really restrict it.
The observed fact of common descent, nested hierarchies all strongly support evolutionary theory but what do they do for a theory of intelligent designer? Why would one expect ID to explain the evidence found so far that so strongly support common descent?
How can we determine if there is a better explanation for these data to be found in intelligent design?
I'd argue that without more knowledge about the designers, we will be unsuccesful in doing so.
RBH's voiced my concerns quite well
quote:
I'd word that slightly differently, saying something like accounting for these similarities in an ID context without invoking common descent as evolutionary biology understands it requires considerable development of hypotheses about the intentions, methods, and proclivities of the hypothesized designing agency(ies). Those hypotheses require a design theory that is considerably more clearly articulated than anything I have ever read in the ID literature.
So far science has proposed the best explanation for the observations given our knowledge.
[ 25. December 2002, 21:48: Message edited by: Frances ]
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 25. December 2002 22:24
I thought of some remarks on Hunter's last posting as I picked up Mayer's book and read its title as I did. Hunter wrote quote:
It seems to me that the reasoning for [Argon's claim, denoted as] Claim A must take the form of "the designer could not do it this way." Such a premise, it seems to me, is impossible to prove, so the argument must take a probabilistic form, ie, it is highly unlikely that the designer would do it this way. I call this the "cosmic conspiracy" form of argumentation, ie, yes the designer could have done it this way, but, because it is so unlikely, it would be tantamount to a conspiracy because it would be done in such a way as to make it appear that there was no design involved.
and again quote:
I am still at a loss to understand why God ought not use congruent designs for the sequences. At best, it seems to me, the ID critic can point out that ID has no reason to prefer these sequences over any others (from within the set of viable sequences). That would certainly be valid but comes nowhere close to supporting Claim A.
The difficulty arises because there is nothing in the ID literature that I have read that corresponds to Mayer's book, What Evolution Is. Open the book and one sees a Table of Contents with a Section called "How Are Evolutionary Change and Adaptedness Explained?", with chapters titled "How and Why Does Evolution Take Place?" and "Variational Evolution" and "Natural Selection". Read the latter chapter and one finds text describing mechanisms and variables that the theory holds to explain the phenomena. Intelligent Design has nothing remotely similar. One looks in vain in the ID literature for descriptions of classes of phenomena that do not map directly onto those already described by evolutionary theory, for relevant variables that account for phenomena, and for mechanisms implemented in matter and energy that explain the phenomena.
On a number of occasions I have seen ID proponents ask 'What evidence would lead an ID critic to suspect (or conclude) that intelligent agency was/is involved in the design of living things?' The great difficulty in answering that question arises from the paucity of genuine theory in ID. If ID critics make what to ID proponents seem erroneous or mistaken or unreasonable conjectures about the purported intelligent agency(ies), it is because ID itself tells us nothing about those agencies. Having no hypotheses or even conjectures about the intelligent agency(ies) from the proponents of ID, what is one to do?
I said above that quote:
Scientific theories explain relationships among classes, not individual instances. Individual instances are explained by first identifying the class membership(s) of the instance, and then referring to laws and principles applicable to the class. The account will also provide criteria that allow us to classify and explain new instances of phenomena, instances not heretofore observed. Finally, the account will provide guidance concerning where and how to look for new phenomena and classes: it will generate a research program.
As far as I can tell, Intelligent Design has not reached the very first step, providing a system of classes, with criteria for assigning instances to classes. As Wesley Elsberry has remarked, so far in published material Dembski's Explanatory Filter has been applied (at least partially) to just four instances, only one of them biological: the Caputo case, the Contact movie prime numbers case, Dawkins' METHINKS IT IS LIKE A WEASEL case, and (in part) the E coli flagellum. That's a very small sample from which to begin to develop classes and hypotheses about relationships among classes. Absent that, absent a theory of ID that provides something like a testable hypothesis or two, ID critics have no option but to imagine what an intelligent agency (or agencies) might or might not do. Hunter wrote quote:
I agree that more articulation of ID will be helpful; however, it seems to me that your implying an additional burden on ID (as opposed to evolution) that doesn't seem warranted. As I have discussed there are plenty of anomalies from the evolutionary perspective and evolution has tremendous wiggle room given their use of the random-design null hypothesis.
To this point ID has borne virtually no burden, say nothing of an "additional" burden. Until there is a clear articulation of ID as a theoretical system, with classes that group phenomena, relevant variables identified, and proposed relationships expressed in testable hypotheses, ID has no particular call to complain about the musings someone else might entertain about the designer(s) and its(their) intentions, methods, and mechanisms.
If we set all that aside, Hunter raised an interesting and pertinent point in the OP: When (or under what circumstances) does the invocation of "sub-theories" serve to conceal a genuine theoretical difficulty? Are there genuine anomalies so serious that evolutionary theory is clearly inadequate and ought be replaced, and if so, by what? What alternative developed theory is there that accounts for the data that evolutionary theory successfully accounts for as well as accounting for the (alleged) anomalies that evolutionary theory cannot? Is it that current evolutionary theory is inadequate, or merely incomplete? Are the "sub-theories" merely special pleading to save a fatally flawed theory, or are they extensions and elaborations of a theoretical framework that provides coherent explanations and fruitful research directions? How can one tell the difference?
RBH
IP: Logged
|
|
Cornelius G. Hunter
Member
Member # 81
|
posted 25. December 2002 23:02
Frances:
You quoted me out of context in your attempt to present ID as an ambiguous theory. I wrote:
quote:
Obviously designed objects will contain many correlated characters, ie, they will produce high congruence. [and then later] Let us say that in 20 years, after our knowledge of cellular operations has increased tremendously, we conclude that protein sequences are indeed arbitrary. I am still at a loss to understand why God ought not use congruent designs for the sequences.
These two statements are not at all opposed. The former is in reference to character sets that are correlated across designed objects (I gave an aircraft example in an earlier post); the latter is in reference specifically to protein sequences, which may or may not be correlated in this way.
In the remainder of your post you appear to be presupposing that common descent is true:
quote:
The observed fact of common descent, nested hierarchies all strongly support evolutionary theory.
In any case, you later state that the evidence "so strongly" supports common descent. I have studied the evidence carefully and arrived at precisely the opposite conclusion. I believe you will have a difficult time supporting these claims.
I agree with you that because God is sovereign we will always be unable to model how He goes about designing in any sort of exacting sense. It is never going to be like Newtonian physics. Therefore, IMHO, ID will not be able to escape some degree of an ad-hoc nature. This, of course, has no bearing on the truth status of ID (unless you assume naturalism in which case you're rejected ID from the very start anyway).
On the other hand, we have centuries of engineering theory and practice accrued with which to evaluate, learn from, and predict design in biology. A common criticism of ID is that it fails to provide a research framework and questions for further research. I would strongly disagree with this. Just as a simple example, in our discussion in this thread I have noted that I would expect the variations in protein sequences to be there for a reason whereas evolution traditionally assumes no function unless one is obvious. In this case, it is ID and not evolution that provides the research motivation.
Also, I should add that if ID is to some degree ad-hoc, this does not put it at a disadvantage compared to evolution.
--Cornelius
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 26. December 2002 00:56
Hi Cornelius,
I am glad to hear that you have studied the vaste amounts of evidence supporting common descent. I understand that you disagree with the conclusions reached based on these data. I believe though that you will have a difficult time supporting these claims. The fact of common descent is strongly rooted in a variety of independent evidences.
I did not intend to 'quote you out of context' I was merely pointing out that your arguments seem to be arbitrary in that character sets may or may not be correlated across designed objects. In fact I doubt that one could find nested phylogonies for most designed objects.
I am also glad to hear that you consider the designer to be God. It is refreshening to see IDists be clear and upfront about their intelligent designer. Indeed, we may never be able to understand God and how He created but He surely gave us some strong evidence that indicates that the theory of evolution may not be that far off.
Applying engineering concepts to biology may be helpful in our understanding of their principles but one should not confuse the analogy beyond the realm of its applicability. ID may provide for research ideas because it inherently doubts any natural mechanisms but that does not mean that ID adds to scientific inquiry. I doubt that ID would contribute to scientific inquiry beyond where science would go itself. To suggest that ID furthers research merely limits the contributions of ID to doubting evolution. But science itself doubts itself and I would claim that science through its doubts of evolutionary mechanisms has done far more for our understanding so far. Endosymbiosis or genetic toolboxes come to mind for instance as does neutral theory.
ID is not only ad hoc so far but it is far more disadvantaged than evolution since it lacks mechanisms, processes, simulations etc. So far there is no competition between evolutionary science and ID. ID so far seems to be limited to showing that evolution could not explain certain features but ID has failed to provide for mechanisms and processes to such an extent that even Dembski seems to argue now that ID should not be held to such standards.
Of course the truth status of ID has no relationship to our faith interpretations of His word either. We should be careful not to mix our faith and science, since both will suffer.
I appreciate your forwardness in addressing some of my statements. They are very helpful to me in understanding your position. As a Christian I have looked at the same evidence or perhaps additional evidence that you have and found that science has done a great job so far discovering how God 'created'.
In Christ
Francis
IP: Logged
|
|
Cornelius G. Hunter
Member
Member # 81
|
posted 26. December 2002 01:51
RBH:
Good posts. I'd like to respond to your most recent one as I fairly agree with your earlier one. You made the point that ID lacks any sort of explicit theoretical foundation of mechanism which evolution seems to have:
quote:
One looks in vain in the ID literature for descriptions of classes of phenomena that do not map directly onto those already described by evolutionary theory, for relevant variables that account for phenomena, and for mechanisms implemented in matter and energy that explain the phenomena. --RBH
I trust you are aware that this statement would be equally valid with "ID" and "evolutionary theory" swapped (with one exception discussed in next paragraph). After all, design predates evolution. A common complaint against design theory is that it is not published in peer reviewed journals. I would say probably the majority of life science journal articles entail design theory. From the Krebs cycle to ATSase, molecular biology has been investigating the design of cellular machinery.
Is ID required to meet the standards of naturalism
The one exception in your statement is in the final clause, ie, "mechanisms implemented in matter and energy." It is probably the fundamental tenet of evolutionary theory that it be restricted to naturalistic explanations. Everything must be described in terms of, as you say, mechanisms implemented in matter and energy. The fundamental tenet of ID, on the other hand, is that this is insufficient. There must be a designer who is sovereign. Hence, it is not clear why one would expect such a mapping in the first place.
Why would one expect ID to provide "mechanisms implemented in matter and energy" which are alternatives to evolution? Indeed, I would argue that if it were to do so then it would lose its distinction and would be a subtheory of evolution. And obviously, it would violate the very premise for its existence, ie, that such explanations are insufficient. The point is that ID is a different type of scientific theory. Evolution attempts to explain things in terms of naturalistic mechanisms, ID does not. It doesn't make sense to apply the standards of naturalism to ID.
Evolution is speculative
As an aside, I would note there is a great deal of speculation in evolution's explanations. So while there may be long books on how evolution is supposed to have worked, it is not clear that how compelling the explanation is. Yes, evolution is naturalistic, there is no doubt about this. And hence it can construct a great many explanations for how this or that mechanism could have worked. But the details are often lacking and, in fact, there is little evidence that the mechanism actually did do what we are being told. How did the DNA code and associated machinery evolve? There are dozens of naturalistic explanations that seem no better than Descartes' explanation that the solar system was created by vortices. The fossils do not appear as expected, so we have punctuated equilibrium. The origin of the cell is difficult to explain, so we have the "RNA world." Naturalistic explanations, yes, but highly speculative. Again, your sentiment regarding ID applies just as well to evolution:
quote:
If ID critics make what to ID proponents seem erroneous or mistaken or unreasonable conjectures about the purported intelligent agency(ies), it is because ID itself tells us nothing about those agencies. --RBH
I don't mean to say evolution fails to explain itself -- it does. But those explanations are so flexible I believe Popper was correct about the theory's strong resistance to falsification.
How to falsify ID
You wondered whether ID is falsifiable:
quote:
Having no hypotheses or even conjectures about the intelligent agency(ies) from the proponents of ID, what is one to do? --RBH
Ironically, this leads to a major problem for evolution, not ID. First, there is, I think, nothing mysterious about how to falsify design. In fact, due to Darwin it is deeply imbedded in evolutionary thought. It is a three pronged argument:
1) Show that the designer's actions make little sense,
2) Show that naturalistic mechanisms are sufficient to explain the origin of species,
3) Show that the preponderance of scientific evidence/analysis strongly points to evolution.
Any of these is sufficient to falsify ID, or at least effectively falsify it by rendering ID redundant. Darwin and evolutionists attempt (and fail) to do all three. Your objection above, of course, applies to the first part, ie, showing that the designer's actions make little sense. How can this be done with no model of the designer? Actually, it is not so difficult. As Darwin showed, you go about it a bit differently. Instead of trying first to model the designer, you merely look at biology, make deductions about what the designer must have done, and argue it makes no sense. You never have to give a complete description of the designer.
Unfortunately for evolution, the required designer's actions are not quite so obviously nonsensical. In fact, what evolutionists end up doing is making rather questionable arguments in the 1st category. Questionable because they rely on particular concepts of God which are rather controversial amongst believers. In other words, Darwin and evolutionists make, what amounts to, religious claims about God's actions which are by no means derived from the Scriptures. Here are some typical quotes:
quote:
Orchids manufacture their intricate devices from the common components of ordinary flowers, parts usually fitted for very different functions. If God had designed a beautiful machine to reflect his wisdom and power, surely he would not have used a collection of parts generally fashioned for other purposes. Orchids were not made by an ideal engineer; they are jury-rigged from a limited set of available components. Thus, they must have evolved from ordinary flowers. [Stephen Jay Gould, "The Panda's Thumb," in The Panda's Thumb, p. 20, W. W. Norton and Company, New York, 1980]
quote:
Odd arrangements and funny solutions are the proof of evolution-paths that a sensible God would never tread but that a natural process, constrained by history, follows perforce. No one understood this better than Darwin. Ernst Mayr has shown how Darwin, in defending evolution, consistently turned to organic parts and geographic distributions that make the least sense. [Stephen Jay Gould, "The Panda's Thumb," in The Panda's Thumb, pp. 20-21, W. W. Norton and Company, New York, 1980.]
quote:
The point is that not only can natural selection account for the design of organisms, but it amounts to blasphemy to attribute it to God's special action. Consider the human jaw. We have too many teeth for the jaw's size, so that wisdom teeth need to be removed and orthodontists make a decent living straightening the others. Would we want to blame God for such defective design? A human engineer could have done better. Evolution gives a good account of this imperfection. Brain size increased over time in our ancestors, and the remodeling of the skull to fit the larger brain entailed a reduction of the jaw. Evolution responds to the organisms needs through natural selection, not by optimal design, but by tinkering as it were, by slowly modifying existing structures. Consider now the birth canal of women, much too narrow for easy passage of the infant's head, so that thousands upon thousands of babies die during delivery. Surely we don't want to blame God for this defective design or for the children's deaths. Science makes it understandable, a consequence of the evolutionary enlargement of our brain. F. Ayala, http://www.kcfs.org/ayala.html]
As you can sense from the sentiment expressed, these are the strong arguments for evolution. They have to be because evolution fails on the second and third types of arguments. But you can also see the sentiment expressed in these quotes is fairly metaphysically laden. So this is what I meant when I said the 1st type of argument leads to problems for evolution. Yes, it is an extremely powerful argument, but clearly it is making some rather heroic assumptions about God and therefore makes evolution dependent on controversial religious claims.
What is ID research?
You wondered if ID has really done much of any research thus far:
quote:
To this point ID has borne virtually no burden, say nothing of an "additional" burden. Until there is a clear articulation of ID as a theoretical system, with classes that group phenomena, relevant variables identified, and proposed relationships expressed in testable hypotheses … --RBH
I do believe that there is room for work to be done along the lines of what we might call "machine design theory." Given a set of machines that must operate under certain constraints (natural laws, environments, etc.) and some degree of commonality in those constraints (they all are subject to the same laws and some degree of similarity in environment), then what can we say about constraints on their design features?
But I see this as a rather limited, esoteric, and almost tangential line of research for ID. You ask for classes, phenomena, variables and so forth. That is the bread and butter of ID research and has been going on as long as biology. But because biology is not like physics those relationships are complicated. Physics has a relatively small number of laws, biology is more like a tapestry. Yes, we can argue it ultimately boils down to the actions of those laws with matter. But the systems are so complex that the manifestations are wildly varying. Yet they are worth studying because generalizations at the "manifestation" level are fruitful, to the extent they are possible. This is the research of ID, not fulfilling some required modeling of the designer as envisioned by those committed to naturalism.
Another line of research, more epistemological in nature, is to provide results that support its right to exist. In other words, argue for ID's veracity. What this boils down to is arguing for ID's premise that naturalistic explanations are insufficient. From a scientific perspective this is a rather simple task. Evolution provides an excellent platform from which to show this. We need not contrive naturalistic explanations, evolutionists have already done this. Clearly evolution is our best shot at explaining biology naturalistically. We need merely point out the weaknesses and internal contradictions within evolution. Unfortunately, there is much more to this than mere science, philosophy and history. There are religious and political motivations involved which quickly blur the picture, but that is another story.
--Cornelius
IP: Logged
|
|
Cornelius G. Hunter
Member
Member # 81
|
posted 26. December 2002 02:29
Francis:
I am overjoyed to hear that you are a believer in our Lord and Savior Jesus Christ. I suspect we will agree to disagree on common descent, but for the record, when you say:
quote:
we may never be able to understand God and how He created but He surely gave us some strong evidence that indicates that the theory of evolution may not be that far off. -- Francis
I can't imagine what evidences you are referring to since all the evidences that evolutionists use are ambiguous or in fact flawed. I'm also having difficulty with your claim that:
quote:
In fact I doubt that one could find nested phylogenies for most designed objects. -- Francis
Even evolutionists (at least the ones I've discussed this with as well as Jukes and Berra in print) agree that designed objects contain correlated characters across the different objects. Did my aircraft example not make sense to you?
Also, you wrote:
quote:
ID may provide for research ideas because it inherently doubts any natural mechanisms but that does not mean that ID adds to scientific inquiry. I doubt that ID would contribute to scientific inquiry beyond where science would go itself. To suggest that ID furthers research merely limits the contributions of ID to doubting evolution. -- Francis
I do not believe that you are doing justice to ID here. Yes, ID states that natural mechanism is insufficient, but this is not the driving force behind its research. Things are intelligently designed so we look for rational design. Where evolution looks for happenstance and vestiges, ID looks for function. God is sovereign so we cannot model His creative actions, but we can think His thoughts after Him, and follow his tracks, so to speak. And yes, we will often be surprised. Is the DNA code a triplet or what? We investigate and find out. As we learn, we continue to ask questions and pursue research.
I think you are seeing ID in the context of evolution rather than as a different paradigm. For example, you write:
quote:
ID is not only ad hoc so far but it is far more disadvantaged than evolution since it lacks mechanisms, processes, simulations etc. -- Francis
Again, this I don’t think this is fair to ID. Every mechanisms, processes, simulation dealing with the operation of organisms is accessible to ID (indeed, I would say often encouraged by design thinking). Why would an molecular dynamics simulation not fit within ID?
Finally, I wonder about this sentiment you expressed:
quote:
We should be careful not to mix our faith and science, since both will suffer. -- Francis
This obviously does not derive from science nor the Scriptures, so I'm not sure why you say this.
Respectfully, Cornelius
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 26. December 2002 03:23
Perhaps the reason that Hunter finds the evidence for common descent weak is that he misunderstands crucial points.
E.g., he has repeatedly alleged, without evidence, that designed objects will produce nested hierarchies. But it just ain't so:
quote:
source
Although it is trivial to classify anything subjectively in a hierarchical manner, only certain things can be classified objectively in a consistent nested hierarchy. The difference drawn here between "subjective" and "objective" is crucial and requires some elaboration, and it is best illustrated by example. Different models of cars certainly could be classified hierarchically - perhaps one could classify cars first by color, then within each color by number of wheels, then within each wheel number by manufacturer, etc. However, another individual may classify the same cars first by manufacturer, then by size, then by year, then by color, etc. The particular classification scheme chosen for the cars is subjective. In contrast, human languages, which have common ancestors and are derived by descent with modification, generally can be classified in objective nested hierarchies (Pei 1949; Ringe 1999). Nobody would reasonably argue that Spanish should be categorized with German instead of with Portugese. The difference between classifying cars and classifying languages lies in the fact that, with cars, certain characters (for example, color or manufacturer) must be considered more important than other characters in order for the classification to work. Which types of car characters are more important depends upon the personal preference of the individual who is performing the classification. In other words, certain types of characters must be weighted subjectively in order to classify cars in nested hierarchies; cars do not fall into natural, unique, objective nested hierarchies.
Because of these facts, a cladistic analysis of cars will not produce a unique, consistent, well-supported tree that displays nested hierarchies. A cladistic analysis of cars (or, alternatively, a cladistic analysis of imaginary organisms with randomly assigned characters) will of course result in a phylogeny, but there will be a very large number of other phylogenies, many of them with very different topologies, that are as well-supported by the same data. In contrast, a cladistic analysis of organisms or languages will generally result in a well-supported nested hierarchy, without arbitrarily weighting certain characters (Ringe 1999). Cladistic analysis of a true genealogical process produces one or relatively few phylogenetic trees that are much more well-supported by the data than the other possible trees.
The degree to which a given phylogeny displays a unique, well-supported, objective nested hierarchy can be rigorously quantified. Several different statistical tests have been developed for determining whether a phylogeny has a subjective or objective nested hierarchy, or whether a given nested hierarchy could have been generated by a chance process instead of a genealogical process (Swofford 1996, p. 504). These tests measure the degree of "cladistic hierarchical structure" (also known as the "phylogenetic signal") in a phylogeny, and phylogenies based upon true genealogical processes give high values of hierarchical structure, whereas subjective phylogenies that have only apparent hierarchical structure (like a phylogeny of cars, for example) give low values (Archie 1989; Faith and Cranston 1991; Farris 1989; Felsenstein 1985; Hillis 1991; Hillis and Huelsenbeck 1992; Huelsenbeck et al. 2001; Klassen et al. 1991).
He also severely misunderstands convergence. Convergence can only produce functionally-relevant similarities, because that is all that selection can "see". Homologies, i.e. similarities between systems that are not necessary for functional similarity between systems, are what allows paleontologists to easily distinguish between these placental wolf and marsupial "wolf" skulls that cre8tionist posted in another thread:
I invite readers to go to The Thylacine Museum and look at the side-by-side comparison of 'wolf' skulls (with cool magnifier lense).
The caption reads:
quote:
Portrayed here are side-by-side images demonstrating the anatomical differences between the skulls of the Grey wolf (Canis lupus) and the thylacine (Thylacinus cynocephalus). In the dorsal view, note that the thylacine has a much broader forehead than the wolf, and there are differences in the design of the zygomatic arches and brain case. Also, the rostrum (snout) of the thylacine is far narrower than that of the wolf, and the thylacine has proportionately larger eye sockets which are rather more square in shape. In the ventral view, one can easily see the great differences in dentition that readily distinguish the two species as being members of distinct mammal groups. The dentition of both species will be represented in greater detail on the following page. Also visible in the ventral view is the thylacine's maxillary palatal vacuity (the two parallel openings in the roof of the mouth). This is a feature that the wolf and other placental mammals do not have.
...on the next page...
quote:
Here I show some diagrams which I have prepared to illustrate the extreme difference in dental anatomy which exists between the thylacine and its placental counterpart, the wolf. The images are portrayed at life size. Although there are also a number of notable differences in post cranial skeletal structure between the thylacine and wolf, I felt that the dentition represented one of the most striking dissimilarities. As you can easily see in the image of the maxilla, the thylacine has 8 top incisors, whereas the wolf has only 6. In the mandible however, the thylacine and wolf have an equal number of incisors. Another major difference is the presence of a specialized shearing tooth, the carnassial, in the wolf. This tooth design is a trademark of the wolf and other members of the placental mammal family Carnivora. Also make note that unlike the wolf, the thylacine lacks large grinding surfaces on its molars. Altogether, the wolf has a complement of 42 teeth, and the thylacine 46.
I can't post the images here because they are copyright protected, but the differences in the tooth-numbering are dramatic.
All commonly-sighted cases of "uncanny convergence" in biology turn out, on investigation, to be externally impressive but superficial when you get down to details.
This has been pointed out many times over the years, so I'm not sure why these cases still get seriously cited.
yersinia
PS: There is also the interesting question of:
If the hypothesized IDer decided that there needed to be some carnivorous canine-type critters in Australia, why bother with all the genetic engineering that would be required, when a simple aboriginal boat sufficed to bring dingos to Australia only ~15,000 years ago?
Such ID puzzles are absolutely ubiquitous in biogeography. To me they indicate strongly that whatever creativity made these wonderful adaptations was, for some odd reason, highly constrained so that "design information" could not be transmitted across deep water barriers and instead had to be re-invented from scratch each time the adaptation was "needed" in particular locations, by modification of whatever happened to be available. Strangely, such geographical constraints did not apply to flying birds, sea mammals, and other easily-dispersed organisms.
If you can find an ID theory that can explain this (and "the designer's actions are mysterious" is not an explanation), I'll eat my hat. If on the other hand you give natural selection the credit for these instances of creativity, then I guess natural selection can "design" things after all, and quite skillfully too...
IP: Logged
|
|
Cornelius G. Hunter
Member
Member # 81
|
posted 26. December 2002 11:48
Yersenia:
You argue that designed objects will not produce nested hierarchies. I already gave the aircraft example, let's consider the example a bit further. Consider all machines that use gasoline as a fuel. Within that set you have various subsets, such as those that move and do not move. Of those that move, some fly, others move along the surface of the earth. In the latter, you have various numbers of wheels, such as 2, 4, 6, 8, …, and a few outliers with odd number of wheels. Of those with 4 wheels, you have different ratios of interior volume to overall size (eg, vans have a higher ratio). Of those with lower ratios you have different carburetors (fuel injection for sports cars), and so forth. I am contriving this example off the top of my head, but perhaps you can explain why this cannot be a case of a nested hierarchy.
In support of your claim you pasted a few paragraphs from a web document which does not support your claim. The web document is discussing the relationships, across designed objects, of characters which have no influence on performance or are constant over the entire set. What we have called in this thread "arbitrary design decisions."
You seem to be extrapolating from the discussion given there on arbitrary design decisions to conclude that designed objects cannot produce nested hierarchies. Perhaps I am misunderstanding you.
Is there more than one cost function?
At this point it is worth repeating what I wrote earlier about the design cost function potentially being richer than just performance. I was reminded of this because the web document pasted in uses car color which is the same example I used. Car color is clearly not an arbitrary design decision, but it is not a factor in performance. In fact, in this example, car color is correlated with other design decisions, some of which will effect performance.
Evolution considers only performance, and even then only certain aspects of performance. Hence, evolutionists have difficulty seeing beyond performance. I do not mean this as a pejorative comment, just an observation, for I have often seen them hold the design theory to this cost function, as though it is a universal cost function.
Can homologies arise from different genes?
You next go on to discuss homologies, but I'm not sure what the point is. You state that paleontologists can easily distinguish between the placental wolf and marsupial wolf skulls, as though I had stated otherwise. Of course they can, they can also easily distinguish between the bat's wing and human hand, but this does not prevent the pentadactyl pattern from being claimed as evidence for evolution. You pasted a figure of the two skulls which appear highly similar yet are supposed to have evolved independently. You say the similarities are "superficial." I have heard this said many times before, but how is it that these similarities are superficial whereas homologies such as the pentadactyl pattern, which exhibit a large degree of variance (compare the porpoise, bat and horse) are significant? We should also note that homologies can develop from different genes, or otherwise follow different development patterns.
An interesting challenge?
Finally, you issued a challenge which sounded interesting but, forgive me if I am slow this morning, I had trouble following. You wrote:
quote:
If the hypothesized IDer decided that there needed to be some carnivorous canine-type critters in Australia, why bother with all the genetic engineering that would be required, when a simple aboriginal boat sufficed to bring dingos to Australia only ~15,000 years ago? -- Yersinia
Can you elaborate a bit?
--Cornelius
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 26. December 2002 15:15
quote:
You argue that designed objects will not produce nested hierarchies. I already gave the aircraft example, let's consider the example a bit further. Consider all machines that use gasoline as a fuel. Within that set you have various subsets, such as those that move and do not move. Of those that move, some fly, others move along the surface of the earth. In the latter, you have various numbers of wheels, such as 2, 4, 6, 8, …, and a few outliers with odd number of wheels. Of those with 4 wheels, you have different ratios of interior volume to overall size (eg, vans have a higher ratio). Of those with lower ratios you have different carburetors (fuel injection for sports cars), and so forth. I am contriving this example off the top of my head, but perhaps you can explain why this cannot be a case of a nested hierarchy. In support of your claim you pasted a few paragraphs from a web document which does not support your claim. The web document is discussing the relationships, across designed objects, of characters which have no influence on performance or are constant over the entire set. What we have called in this thread "arbitrary design decisions."
Did you even read the quote? The very first sentence pointed out that anything can be subjectively classified into a nested hierarchy if you arbitrarily pick characters. This is exactly what you do above. The point is that your "tree" would not be produced by an analysis of other subsystems of gasoline-driven machines, e.g. tires, liscense plates, GPS units, radios, onboard computers, whatever. On the other hand, in biology there are a large number of systems (genes, limbs, skulls, etc.) that produce highly-congruent nested trees. Other fairly similar examples are things like languages and scribe-copied documents, both of which are produced by a process of copying and gradual modification (although in these cases the possibility of lateral transfer is somewhat higher than it is in eukaryote biology).
As for web references, if they cite the primary literature then you either have to show they are mis-using the literature, or that the literature itself is wrong. Theobald cites a large number of papers discussing the difference between arbitrary and natural hierarchies -- designed objects like cars and planes produce the former, copied & gradually modified objects (like languages, scribe-copied documents, and...organisms) produce the latter.
I'll include some of Theobald's refs so that interested parties can look them up:
Archie, J. W. (1989) "A randomization test for phylogenetic information in systematic data." Systematic Zoology 38: 219-252.
Faith, D. P., and Cranston, P. S. (1991) "Could a cladogram this short have arisen by chance alone?: on permutation tests for cladistic structure." Cladistics 7: 1-28.
Farris, J. S. (1989) "The retention index and the rescaled consistency index." Cladistics 5:417-419.
Felsenstein, J. (1985) "Confidence limits on phylogenies: an approach using the bootstrap." Evolution 39: 783-791.
Hillis, D. M. (1991) "Discriminating between phylogenetic signal and random noise in DNA sequences." In Phylogenetic analysis of DNA sequences. pp. 278-294 M. M. Miyamoto and J. Cracraft, eds. New York: Oxford University Press.
Hillis, D. M., and Huelsenbeck, J. P. (1992) "Signal, noise, and reliability in molecular phylogenetic analyses." Journal of Heredity 83: 189-195. PubMed
Ringe, D. (1999) "Language classification: scientific and unscientific methods." in The Human Inheritance, ed. B. Sykes. Oxford: Oxford University Press, pp. 45-74.
quote:
You seem to be extrapolating from the discussion given there on arbitrary design decisions to conclude that designed objects cannot produce nested hierarchies. Perhaps I am misunderstanding you.
Of course designed objects can produce just about anything, because a hypothetical designer can always be invented who wants to produce X for goodness-knows-what reason. This is a major problem for ID "theory", no predictions are made unless some specifications are put on the hypothetical IDer, and no one wants to even hypothesize any such specifications (you don't have to have foreknowledge of the designer, just a hypothesis...this is how science proceeds).
But you said that ID predicts congruent phylogenies. I am arguing that this is not established or even likely based on what we know about designed objects.
quote:
Is there more than one cost function?
This section seems like you are trying to say something about how the designer would design things so that congruent phylogenies resulted due to functional constraints, or something. But what you have to explain, in order to explain things as well as current theory, is how all of those arbitrary characters (many of them, such as DNA degeneracy, absolutely known to be functionless differences) produce statistically the same nested hierarchical trees! If you can't do that then there's no reason to switch from the current explanation.
quote:
Can homologies arise from different genes?
You next go on to discuss homologies, but I'm not sure what the point is. You state that paleontologists can easily distinguish between the placental wolf and marsupial wolf skulls, as though I had stated otherwise. Of course they can, they can also easily distinguish between the bat's wing and human hand, but this does not prevent the pentadactyl pattern from being claimed as evidence for evolution. You pasted a figure of the two skulls which appear highly similar yet are supposed to have evolved independently. You say the similarities are "superficial." I have heard this said many times before, but how is it that these similarities are superficial whereas homologies such as the pentadactyl pattern, which exhibit a large degree of variance (compare the porpoise, bat and horse) are significant?
Because the homologies all correlate with each other to a high degree of statistical confidence, producing a Linnean-type classification, whereas those features that you would expect would be the important features (as revealed by you example of classification of gas-driven machines based on function, or John Bracht's imaginings that amino acid sequence won't turn out to be largely degenerate with respect to structure and function after all) in fact don't correlate with the Linnean-type classification.
If the genes and proteins of penguins, sharks, dolphins, seals, etc. grouped together, and bats and birds grouped together, etc., then you'd have an argument, but they don't. This is a massive mystery from an ID perspective but easily explained by evolution.
quote:
We should also note that homologies can develop from different genes, or otherwise follow different development patterns.
| | |
