|
Author
|
Topic: Evolving Inventions
|
ASCSCommanding
Member
Member # 682
|
posted 07. March 2003 14:57
Frances,
Thanks for the comment. Sorry for taking so long to get back. Although I’d like to participate more, it looks like a few posts a week will be limit for a while.
From reading John’s paper, I conclude that identifying the hypervolume in which an object is found is not easy and depends upon the underlying system that causes the object to exist. If I understand John’s discussion in his paper of the two versions of Dawkin’s biomorph program (the 9 and 16 gene versions) the exact same biomorph could be in a hypervolume of either dimension 9 or dimension 16, depending upon which version you’re running. Therefore, the dimensionality of the hypervolume is not a property of only the object itself. Furthermore, a hypervolume, by definition, cannot be exited by any trial-and-error method available, so the hypervolume an object exists in contains all objects that can be reached by a trial-and-error method. It seems to me that John must intend some limitation on what trial-and-error methods can be used, but I’m not sure how to describe those limitations. This is my understanding, so please correct me, John, if I’m wrong. The upshot is that all the solutions reached by any of your GAs are all within one hypervolume. However, this also makes it difficult to clearly identify that two objects are in different hypervolumes. (We can clearly show that they are in the same hypervolume, but showing that they are in separate hypervolumes is harder.) John also is claiming that if two objects exist in separate hypervolumes then TRIZ-like inventiveness is needed to get from one to the other.
Now, I would expect that any argument against the ToE based on TRIZ would try and either show clearly that separate hypervolumes exist, or show that inventiveness has occurred. Given the above paragraph, your long running discussion with John, plus my earlier claim that all possible genomes are reachable by the observed changes that occur to genomes when they reproduce (one can imagine genomes and ways in which they are altered which would put different genomes in clearly distinct hypervolumes) together indicate that identifying clear hypervolumes is problematic.
My main point has been in response to John claiming to detect inventiveness. In particular, one line of investigation that he should pursue is to ask how much information do we need to have about precursors of an object (biological or otherwise) to clearly say that the object shows inventiveness. It seems to me that an object might or might not show inventiveness depending upon what preceded it. It seems likely that the TRIZ studies would already provide most of the information needed to answer this question. It would be problematic to John’s claim if the amount of information needed was much greater than that available for the cases he sites as showing inventiveness, likewise it would support his case if the information needed was always less than that available.
If we can’t clearly identify separate hypervolumes and we can’t clearly identify inventiveness one distinct possibility that must be considered is that there is exactly one hypervolume and has been no inventiveness in the history of life. Understand too that I’m using inventiveness here to mean TRIZ-like inventiveness. This may be very different from common usage of inventiveness. No matter how inventive we might think the results of GAs are, if they are the result of a GA, they are not TRIZ-like inventive.
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 07. March 2003 16:52
ASCSCommanding said:
quote:
If I understand John’s discussion in his paper of the two versions of Dawkin’s biomorph program (the 9 and 16 gene versions) the exact same biomorph could be in a hypervolume of either dimension 9 or dimension 16, depending upon which version you’re running. Therefore, the dimensionality of the hypervolume is not a property of only the object itself. Furthermore, a hypervolume, by definition, cannot be exited by any trial-and-error method available, so the hypervolume an object exists in contains all objects that can be reached by a trial-and-error method. …
I think that John may have some additional constraints on what is meant by “hypervolume” beyond the dimensionality of the original search space per se. I think that an aspect of reachability by (pseudo-)search state steps must be included (modeling as search, for good or bad).
One point that has still not been addressed is my example of the software that had no limit set on the number of “genes” or other “dimensions”, and adds dimensions when it decides to do so. More “dimensions” or greater search range may depend on the search time, but is not inherently limited in the algorithm. In this case the theoretical dimensionality of the raw search space is effectively infinite. Thus no “inventiveness” could possibly happen no matter what was invented, because all points were reachable within the infinite search space -- it is simply sifting of an ever greater and semi-infinite range of possibilities. See next.
quote:
It seems to me that an object might or might not show inventiveness depending upon what preceded it. It seems likely that the TRIZ studies would already provide most of the information needed to answer this question. It would be problematic to John’s claim if the amount of information needed was much greater than that available for the cases he sites as showing inventiveness, likewise it would support his case if the information needed was always less than that available.
I think this is perceptive.
An example: A state is reached by some sort of direct jump to a greater degree of complexity. But it could, alternately, be reached by a series of small steps. Is the first case “inventive”, while the second not? The solution is exactly the same, the only question was how it was reached.
Because the issue of reachable search space is distinct from the actual dimensionality of the raw search space (which may be semi-infinite as proposed above). And the knowledge of what is reachable may be very difficult to analyze -- especially if it is essentially an infinite raw search space and we must simply judge by run time how large the search range is to proceed.
Then for intelligent inventors, consider that they might only assemble ideas in variational ways wherein ideas are a combination of direct experience and serendipitous thought and experience patterns. In this sense the inventor is limited, and cannot invent beyond some limit of possible “search” space. We must then find that a sudden jump might have been due to a very good sense of making connection, but a serendipitous case of source of the ideas (see my Bell telephone example several pages back).
It is very possible that we associate the “inventive” when we don’t know the pathway of the “information” for the idea, or the “search path” for the algorithm. In this form it shares with the IC concept the interpretation as in the form of an “argument from ignorance”.
quote:
If we can’t clearly identify separate hypervolumes and we can’t clearly identify inventiveness one distinct possibility that must be considered is that there is exactly one hypervolume and has been no inventiveness in the history of life. Understand too that I’m using inventiveness here to mean TRIZ-like inventiveness. This may be very different from common usage of inventiveness. No matter how inventive we might think the results of GAs are, if they are the result of a GA, they are not TRIZ-like inventive.
And if we should put in our definitions that an idea is not “inventive” if the inventor previously got the basic ideas from information gained from serendipitous experience and his/her own ability to recognize the “good idea”, then that idea might not be considered “inventive” in this specialized definition -- and thus there might not be any TRIZ inventive ideas generated by humans (at least they would be extremely highly improbable).
[ 07. March 2003, 18:57: Message edited by: gedanken ]
IP: Logged
|
|
Aardvark
Member
Member # 141
|
posted 08. March 2003 05:46
Hi again. A good 3 weeks ago when I was last on ISCID, Yersinia wrote a courteous response to me regarding the alleged jawbone to earbone transition. He claimed that the myth of the earbone transition is nothing but a myth in itself.
I'm not going to give a full critique of the whole story of the alleged reptile-mammal transition right now. Mainly, the whole transition isn't what we are talking about here anyways, and plus it would take a lot of time to write at this time. Suffice to say, however, that there are some very large morphological jumps in the story, and a whole lot of reversals that make the story pretty messy and inelegant, but, to address the point, in response to Yersinia, I do not revise my original thesis: there is no record detailing of the alleged actual jawbone -- > earbone transition itself.
The diagram posted from the T.O. common descent FAQ, as far as I can see does not establish that the jawbone -- > earbone transition we are discussing here took place. There is a reason why Carroll says, in 1988,
quote:
"It is not yet certain when the malleus and incus became incorporated into the middle ear, but the grooves on the medial surface of the dentary that indicate their position of attachment in early Jurassic mammals are missing in Upper Jurassic genera.”
(Carroll, Robert L. 1988. Vertebrate Paleontology and Evolution. W. H. Freeman. New York.)
In the T.O. diagram, from Kardong's textbook on comparative vertebrate anatomy, Morganucodon is cited as the first primitive mammal. Prior to that is the "advanced cynocodont."
This story is somewhat complex. I'm going to do my best to explain it, but it's going to require a look at Kardong's diagram again.
According to the story, the Angular (An) turned into the tympanic annulus (Ty)--a ring of bone supporting the tympanic membrane, commonly called the eardrum. The Articular (Ar), which was part of the reptilian quadrate-articular ("articulate" means "to unite by a joint") jaw-joint, turns into the Malleus (Ma), the "hammer" bone in the middle ear. The quadrate becomes the Incus (or anvil). The Stapes (S), pronounced stapees, is just about the only bone that always has and always will be the same name: the Stapes (or stirrup).
The Talk Origins Macroevolution FAC (fecklessly asserted claims) cites 2 papers to say when exactly this transition supposedly takes place. We'll call them Wang et al. (2001) and Luo et. al. 2001.
Let's take a look at Luo first. Luo discusses the discovery of a fossil, Hadrocodium, which appears to be a very early mammal from the late Jurassic (about 195 Ma). It says of this fossil, "It has an enlarged cranial cavity, but no postdentary trough on the mandible, indicating separation of the middle ear bones from the mandible … The absence of these structures indicates that the postdentary bones (“middle ear ossicles”) must have been separated from the mandible."
So, we have no transition here--the earbones are already where they are supposed to be in mammals. This fossil surely tells us that at 195 Ma the jawbone -- > earbone transition had taken place. However it's on the "other side" of the transition and doesn't really help us to know much about it.
Here is what Wang says:
quote:
"In nonmammalian vertebrates with jaws, the craniomandibular joint is between the quadrate region of the palatoquadrate and the articular region of Meckel’s cartilage (or its replacement). In unequivocal mammals, the joint is between the squamosal and the dentary. The definitive mammalian middle ear (DMME) is formed by transference of accessory jaw elements, including the angular, articular plus prearticular, and quadrate, to the cranium of mammals as strictly auditory ossicles (renamed as the tympanic, malleus, and incus). This transference is one of the central topics of comparative anatomy and evolutionary biology of vertebrates."
Translation: In reptiles, the joint between the jaw and the skull is between the quadrate and the articular bones. In mammals, the joint is between the squamosal and the dentary, and the mammal middle ear contains the angular, articular, and quadrate, which become are renamed tympanic annulus, malleus, and incus. In both reptiles, the stapes was already in the ear, as it is in mammals as well.
Wang goes on to say:
quote:
"the only fossil evidence on this critical transference is the presence of persistent grooves on the medial surface of the dentary bone, which may have lodged the anterior end of the postdentary unit (PDU, consisting of the endochondral articular and dermal prearticular, angular, and surangular) in some early mammals."
If I am reading this correctly, then his only evidence that there was a time when the mammalian jaw housed the reptilian post-dentary bones is inferred from a few "grooves" on the dentary bone (the mammalian jawbone). That sounds like severe speculation to me, and is far from conclusive on this point. This fossil too seems to be on the "other side" of the transition: the dentary bone is already doing its job as the mammalian jawbone, and the transitional connection to the reptilian post-dentary bnes is not hard evidence, but is inferred as Wang says it "may have" connected to the reptilian post-dentary bones, and all this inference is based upon a few microscopic "grooves" on the dentary--not hard empirial findings. If that is the "only fossil evidence on this critical transference" then I think that the story transition is beginning to take on mythlike qualities.
Talk origins cites this paper to say, "These new fossil finds clarify exactly when and how the malleus, incus, and angular completely detached from the lower jaw and became solely auditory ear ossicles."
I strongly dissent from this claim. These finds do not tell "how," or "when" and therefore cannot tell "if" the alleged jaw-bone to earbone transition occured. All we really know is that at 195 Ma, there were mammals with a totally mammalian jawbone and earbone. I propose that this Talk Origins claim is a severe over-extrapolation.
In fact, in Luo, the "how" is based upon the ontogeny of a marsupial, where some of the earbones start off big and closer to the jaw, but get seem to get smaller and move away from the jaw. However this is not a selectionist based hypothesis, as the movement of these bones is directional, based upon the genetically preprogrammed ontogeny of the marsupial. Thus, we are not answering the question how Darwinian selection can make functionality jumps--we all know that vast increases in complexity occur during ontogeny, but that does not prove evolution took place, merely that genetics can preprogram large-scale transormations. Otherwise, we'd be arguing that butterfly metamorphoses are evidence of how flying insects evolved from caterpillar-like non-insect precursors.
Going back to Luo et. al., there are 3 fossils cited which are the infamous "double-jaw-hinge" (articular-quadrate AND dentary-squamosal) which allowed the reptile post-dentary unit bones to double as jawbones until the dentary-squamosal took over and allowed the reptile post-dentary bones to magically become a part of the inner ear:
quote:
"The transformation from a more complex “double jaw hinge” (of the articularquadrate and dentary-squamosal in Sinoconodon, Morganucodon, and Haldanodon) to the “single jaw hinge” joint (TMJ) (formed exclusively of the dentary-squamosal in Hadrocodium…"
As noted before, Hadrocodium is the fossil that Luo et all found, the oldest fossil with a true mammal-like jaw, on the mammal side of the alleged transition. Therefore, our jawbone-earbone transition is basically based of the following fossils:
Sinoconodon + Morganucodon + Haldanodon -- > Hadrocodium.
Indeed Kardong's diagram reproduced in Theobald's FAC cites Morganucodon as his "primitive mammal" just before the most advanced mammal skull. This is actually at odds with Luo, who calls Morganucodon a "nonmammalian mammaliaform." However, the reptilian "jawbones" in Sinoconodon, Morganucodon, and Haldanodon are not anything like the mammalian middle-ear--they are part of the jaw, and not part of the ear. It's probably not Kosher to cite a creationist source, but I'll depart from tradition briefly and cite one and then return back to the secular literature. Mehlert, (CRSQ, June 1988) says:
quote:
"Although there is no doubt that Morganucodon and Kuehneotherium both had powerful and standard reptilian jaw joints did they have a contact point between the large dentary and the squamosal and if so, was this an incipiant mammalian type jaw joint? Unfortunately the pitifully small, fragmented and disarticulated fossil material makes this claim unresolvable and any claims are based on inference."
Yet the Talk Origins FAC further claims that, "in Morganucodon, the quadrate (anvil) and the articular (hammer) serve as mammalian-style ear bones and reptilian jaw bones simultaneously." I'm not sure how this can be claimed when, Morganucodon's quadrate and articular are firmily attached to the dentary, and it had only one true earbone--the reptilian stapes.
As Theobald notes, it is possible that Morganucodon could transmit sounds from the dentary, up through the articular to the quadrate, to the stapes, and then to the inner ear, but this is probably how it worked for a lot of reptiles: nothing transitional, and you're still on the "reptile side" of the transition: The stapes is the bone attached to the ear drum. The transition requires a much more radical re-organizing of the parts than just a shifting up of the bones. You may not like Gish, and perhaps Gish has made some mistakes in the past. However, I think he says at least one insightful thing about this problem. Here is what must happen in the transition:
quote:
"This would have required that the stapes (columela) of the reptile become free of its attachment to the ear drum, and the retrarticular process of the articular gained an attachment to the ear drum (because the articular bone of the reptile supposedly became the malleus of the mammal, which is attached to the ear drum). Somehow the reptilian quadrate must gain its freedom, move into the middle ear and insert itself between the stapes and malleus" (Gish, Evolution and the Fossil Record, 1985, pg 100)
In mammals, the bones go as follows:
(http://www.earthlife.net/mammals/images/anatomy/m-ear.gif)
As Theobald illustrates, it is the following in reptiles:
A quick read of Theobald might make one might think that evolutionists have the story all figured out, but even after finding afforementioned incredible, amazing, and groovy dentaries of Repenomamus and Gobiconodon, Wang et. al. still went on to say that, "the most uncertain issue in the evolution of the DMME [Definitive Mammalian Middle Ear] is how the PDU [post dentary unit--includes the articular and quadrate] became detached from the dentary and translocated to the basicranium as ear ossicles."
Translation: "The most puzzling thing about the evolution of the mammalian middle ear, even if these fossils are as great as we are saying they are in this paper, is how the reptile articular and quadrate became separated from the jawbone and found their way in the inner ear."
Hey, wait, isn't this exactly what Gish (and Aardvark) just said? Maybe those creationists aren't as kooky as some people say they are!
Essentially, even if the PDU is attached to the dentary, the main issue is how those bones go from a jaw-location to an ear location. That's the central point that Bracht is trying to make--namely that this requires a drastic reorganization of parts and complexity that Darwinian processes ain't gonna do. The actual transition from jawbones to ear bones is not understood--as Wang makes clear.
Of course Wang takes a stab at it, noting a previously prosed model that increased brain expansion, which supposedly took place in early mammals, naturally increased the distance of the PDU from the jaw. However distancing the PDU from the jaw doesn't therefore take you any closer to the ear. As a matter of fact, it would seem to just take the bones into limbo-land between the jaw and the ear, and would also diminish any hearing advantages that these creatures from having their PDU able to transmit sound from the jaw to the stapes. Furthermore, Wang notes that this model is inconsistent with the new fossil finds, which, incidentally, though supposedly closer to mammals (if one accepts his little groove inferences), have smaller braincases than their alleged precursor, Morganucodon. This led Wang et. al. to say, "This indicates that detachment of the ear ossicles is not necessarily associated with expansion of the brain during mammalian evolution."
Wang proposes his own model basically saying that the PDU got smaller so it could pick up higher frequency sounds, and then detached from the dentary. Of course this is all based upon Wang's soft inference that the PDU was attached to the dentary, but even if it was true, why does the PDU just floating up to the ear, and how does it still transmit sound the way it's supposed to once it is detached from the dentary? Without the "cranial expansion" mechanism, he proposes no reason for the PDU to move its way up to the ear. At this point these bones aren't being used for chewing, and they aren't very good at transmitting sound. Why would they magically go up to the ear? My guess is that they are not being used, and they would be best suited to just disappear, much less would they suddenly move up to the ear and insert themselves. But, more importantly, how do we undergo the re-arrangement of parts that Gish discusses above? There is a big difference between having 2 or 3 little bones down by the jaw doing nothing, and having 3 bones efficiently rearranged inside the middle ear to transmit sounds from the ear drum to the inner ear.
But, as far as fossils are concerned, all this is beside the point. The closest the evolutionists have to the transition is Wang's jaw grooves leading to the inference that a little PDU was sitting there ready to slide up to the ear, and there are no hard fossils showing that PDU in the right place, making the slide, or making the rearrangement.
Carroll (1988) says,
quote:
"extradentary bones were gradually reduced but remained part of the lower jaw even among early mammals [this is because he considers Morganucodon a mammal--this was written long before Wang (2001) made his groove inferences]. The configuration and relative position of these bones indicate without any doubt that the quadrate, articulra, and angular of reptiles are homologous with the incus, malleus, and tympanic of modern mammals
Yet, it is only the position of the bones on either side of the transition allows for this inference which supposedly for Carroll is "without any doubt;" it is not because of a record of the actual transition from jawbones into the earbones.
So, in conclusion, there is not a record transition of the jawbone to the earbone. If there was, it would somehow require that, as Wang proposes, the quadrate and articular become useless for chewing, and then slide up to the ear detaching from the dentary, decreasing their sound transmission to the stapes, and then wedge themselves between the eardrum and the stapes, perhaps one-at-a-time, or perhaps all at once. I don't know which it is, because I've seen no evolutionist go as far as to discuss this necessary part of the transition. But, as said earlier, as far as fossils are concerned, the closest the evolutionists have to getting jawbones into the ears is Wang's jaw grooves leading to the inference that a little PDU was sitting there on the dentary ready to slide up to the ear. There is no record of the journey from the jaw to the ear, and then of the rearrangement of the ear. From what I can tell, the next thing we get is like, Luo's Hadrocodium, which is basicaly normal mammalian bone structure. But, missing fossils aside, theoretically, it's tough to see how this rearrangement can remain functional when you have to detach the stapes from the ear drum.
I think John Bracht has a point to consider: we are seeing high specified complexity in the mammalian middle ear. The alleged evolutionary story has useless bones becoming extremely useful, and it is difficult to see how the intermediates are functional.
Well, I may not be back for another 3 weeks, so when I return I'll see what has become of this!
Thanks to all!
--AA!
[ 08. March 2003, 05:47: Message edited by: Aarvark ]
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 08. March 2003 12:50
Aardvark,
While some of the transitions are not as smooth as we maybe would like them to be, the series nevertheless form a from a paleontological perspective marvelous transitional series. One can of course focus on some of the aspects we do not know, and maybe will never know but that does not diminish the value of the transitional series. So your conclusion that there is no transitional record seems invalid. I find it fascinating to see though how 'rebuttals' seem to focus not on what we know, which of course would support the transitional nature of the fossils, but on what we do not know (yet).
Additionally it seems that embryological evidence supports this transition
quote:
Embryological studies over a hundred years ago demonstrated homologies
between mammalian inner ear bones and “old” jaw bones:
a. stapes homologous to hyomandibular (note this homology also holds for
the columella of amphibians and reptiles with tympanic ears)
b. malleus homologous to articular
c. incus homologous to quadrate
Source
As far as the 'creationist' source on the Morganucadon
quote:
Seventh, concerning Morganucodon, Gish (1995, p. 169) stated, “What is the evidence for a squamosal-dentary joint in these creatures? This evidence consists of an alleged condyle on the dentary.” That statement misrepresents t he data. The condyle is definitely present; it is not “alleged” (Kermack, Mussett, & Rigney, 1973, 1981; Carroll, 1988). The corresponding shallow depression in the squamosal into which the condyle fitted in life corroborates this. Hopson (1987) state d that articulated material has been found with the two bones in contact.
Source
Aardvark complains about the usage of the term "primitive mammal" for Marganucodon and quotes from Luo who refers to it as "nonmammalian mammaliaform". But rather than undermining the original claim it seems to strengthen it.
Also
quote:
(c) Elimination of functional redundancy. For example, it is difficult to hypothesise a direct route by Darwinian evolution from mammalian to reptilian jaws, as they consist of different pairs of bones. However, the fossil intermediates Morganucodon and Kuehneotherium had both quadrate-articular and dentary-squamosal articulation. The following postulated evolutionary sequence from reptilian to mammalian jaws, for which there is considerable fossil evidence, involves selective advantage at each step (Kermack & Kermack, 1984):
Source
Lets look at Luo and the Carnegie Museums
quote:
Previously it was not clear how the accessory mandibular (postdentary) bones of cynodonts (close relatives to mammals) could migrate from the mandible to the cranium in mammals. Luo and Crompton (1994) show that the phylogenetic transformations of several features of the quadrate (incus) made it possible for the postdentary bones to establish new sites of attachment to the cranium.
An interesting paper seems to be "Luo, Z., and A. W. Crompton. 1994. Transformations of the quadrate (incus) through the transition from non-mammalian cynodonts to mammals. Journal of Vertebrate Paleontology 14:341–121."
quote:
Probainognathus is among the earliest known non-mammalian cynodonts with a concave contact facet and a rotated dorsal plate in the quadrate. Thus we hypothesize that it represents a critical step in the phylogenetic transformation that led to the origin of the modern mammalian middle ear and tympanic membrane.
The following Source shows some helpful diagrams for the transformation.
"A new mammaliaform from the Early Jurassic of China and evolution of mammalian characteristics" by Zhe-Xi Luo, Alfred W. Crompton and Ai-Lin Sun.
Let's give an in context quote:
quote:
A fossil from the Early Jurassic (Sinemurian,;195 million years ago) represents a new lineage of mammaliaforms, the extinct groups more closely related to the living mammals than to nonmammaliaform cynodonts. It has an enlarged cranial cavity, but no postdentary trough on the mandible, indicating separation of the middle ear bones from the mandible. This extends the earliest record of these crucial mammalian features by some 45 million years and suggests that separation of the middle ear bones from the mandible and the expanded brain vault could be correlated. It shows that several key mammalian evolutionary innovations in the ear region, the temporomandibular joint, and the brain vault
evolved incrementally through mammaliaform evolution and long before the differentiation of the living mammal groups. With an estimated body weight of only 2 grams, its coexistence with other larger mammaliaforms with similar “triconodont-like” teeth for insectivory within the same fauna suggests a great trophic diversity within the mammaliaform insectivore feeding guild, as inferred from the range of body sizes.
quote:
All other nonmammalian mammaliaforms with small brain vaults (Fig. 3, A to C) have retained the mandibular attachment of the middle ear bones, whereas Hadrocodium and living mammals (Fig. 3, D to I) with larger brain vaults have lost the mandibular attachment to the middle ear.
quote:
The acquisition of “mammalian characteristics” shows an additive pattern in our mammaliaform
phylogeny (Fig. 4C). The transition from the nonmammaliaform cynodonts to living mammals has a stepwise and incremental acquisition of the mammalian characteristics (24–27, 43), and there was no single episode of rapid evolution of a large number of derived characters. The node of crown-group Mammalia [following (42)] has four unambiguous synapomorphies, within the range of three to ten unambiguous synapomorphies for each of the internodal segments on the backbone of the cladogram. This is consistent with a macroevolutionary pattern that prevailed in much of synapsid evolution (24, 25, 43).
All from Luo...
And from the supplementary materials
quote:
No matter which scheme of ordering vs. unordering multi-state characters, all searches consistently show Hadrocodium to be the sister-taxon to the crown-group of Mammalia, or to the clade of (triconodontids + crown-group Mammalia), in all fundamental trees and consensus trees. No matter which scheme of ordering vs. unordering multi-state characters, all searches have consistently placed Hadrocodium closer to the Mammalia than the successively more distant Haldanodon, Morganucodon, and Sinoconodon. Given the matrix presented here, the position of Hadrocodium is stable, so are the successively more distant positions of Haldanodon, Morganucodon, and Sinoconodon.
Now I move on to Wang in more context
quote:
Although developmental studies of extant mammals have long demonstrated homologies of these elements among jawed vertebrates (9, 10), the only fossil evidence on this critical transference is the presence of persistent grooves on the medial surface of the dentary bone, which may have lodged the anterior end of the postdentary unit (PDU, consisting of the endochondral articular and dermal prearticular, angular, and surangular) in some early mammals (3).
quote:
The most uncertain issue in the evolution of the DMME is how the PDU became detached from the dentary and translocated to the basicranium as ear ossicles (3–8).
quote:
Our specimens permit an alternative hypothesis for the origin of the DMME. During evolution of synapsids, the PDU is reduced in size and loosened to enhance hearing of highfrequency airborne sounds (5), whereas the dentary was enlarged for attachment of more muscle to facilitate efficient mastication (3, 32, 33). The position of the OMC in Repenomamus suggests that the common ancestor of mammals probably had a developmental pattern in which Meckel’s cartilage extended from dentary to the ear region. Because of its close relationship with the cartilage, the dentary was probably tilted in position. Reduction of the PDU increasingly weakened its tie to the dentary until a critical point was reached where the dentary, while erecting to a more vertical position during ontogeny, no longer seized the PDU, which was moored at the basicranium by connective tissue.
This hypothesis is similar to the detaching
mechanism of the ear ossicles in marsupials
(6), without requiring brain expansion as the
initial trigger. Modifications in both feeding
and hearing apparatuses toward efficient functions
have led to the decoupling of the PDU and
dentary. Expansion of the brain, along with
changes in the otic capsule, may have caused
displacements of the ear ossicles to a position
either more vertical (6), horizontal (8), or posteriorly distant from the SCMJ (7) in more
advanced mammals.
quote:
Developmental studies of living mammals have revealed that the posterior end of Meckel's cartilage forms the anlage of the malleus and that the middle portion of the cartilage degenerates in the later stages of ontogeny. Its sheath becomes the sphenomandibular ligament (pterygomandibular in monotremes) and the anterior ligament of the malleus (3, 9, 10). The OMC in Repenomamus and Gobiconodon provides evidence for the relationship of Meckel's cartilage with the DMME in early mammals, which is otherwise inferred only from embryological evidence of living mammals. It also shows that, while the anlage of the malleus is reduced, or posteriorly shifted, to form the malleus, a significant middle segment of Meckel's cartilage is persisted and ossified in adults, probably remaining in its early ontogenetic position. A similar condition is probably present in other early mammals, such as triconodontids and symmetrodontids. The function of the OMC in adults of these mammals is unclear. A dorsal facet on the OMC of Repenomamus (Fig. 1F) suggests that muscle was attached. If so, the OMC may have functioned as the inflected angular process in marsupials (24), or the pterygoid shelf in multituberculates (25), for partial insertion of the medial pterygoid muscle that originates on the pterygoid region of the skull. When opening and closing the lower jaw, the attached OMC could rotate with the jaw, with its contact at the lateral flange serving as the fulcrum. Mastication may not have interfered with hearing in Repenomamus.
It seems that the evidence for a transitional series from reptile to mammal, including what happened to the hearing is quite well supported by consistency, paleontological evidence and embryological evidence.
[ 08. March 2003, 14:33: Message edited by: Frances ]
IP: Logged
|
|
John Bracht
Member
Member # 5
|
posted 08. March 2003 16:43
Hi all,
With finals approaching I really can't be spending much time here, but I want to make a few quickie comments. Nelson, thanks for stepping in and making some key points during my absence. Once again, I have great desire to respond to dozens and dozens of arguments I see here, but I just can't because of time.
At any rate, Francis responded to me by asking for clarification of precisely what I mean by various terms like hypervolume and inventive. I find this request very amusing, because Francis has spent probably tens of thousands of words insisting that GA's and the Darwinian process are capable of both englarging the hypervolume of possibilities in which they reside, and producing true inventions. Now, he's saying that he doesn't really understand what these terms mean! Francis, you can't have your cake and eat it too--either you already understand the terms (and have refuted my arguments, as you've insisted over and over again), or you don't understand the terms and you have no business insisting all along that I'm wrong and you have proven so. Since you either understand these terms or you don't (this covers all logical possibilities), you're incorrect in one way or another. It would be nice to see you retract an argument or two. (and you might want to be careful not to make logically contradictory assertions, like saying that you've refuted an argument and then saying that you don't really understand that argument).
As for the specifics of what I'm requesting of you (remember, I have been pressing you to show how the Guo et al example falsifies my arguments by specifically showing enlargement of the hypervolume of possibilities), let me just say that you shouldn't need any fancy definitions. Basically, all I'm asking is that you show that the algorithm output a solution which was not accessible from the initial program. Show how the program reached a solution that was not part of the original possibility space, that is, there was no way for the program to get to the solution it eventually found. To refer back to Dawkin's biomorph program, show (the equivalent of) how a "blue" biomorph could arise from a program which initially had no gene for color. Show how the parameters of the system were adjusted purely by the GA itself to allow the novel solution to be found (a solution that was not originally reachable, in any way, with the initial parameters). If you can show this, I'll be happy to drop my argument that GA's cannot enlarge the hypervolume of possibilities.
Gedankin:
quote:
One point that has still not been addressed is my example of the software that had no limit set on the number of “genes” or other “dimensions”, and adds dimensions when it decides to do so. More “dimensions” or greater search range may depend on the search time, but is not inherently limited in the algorithm. In this case the theoretical dimensionality of the raw search space is effectively infinite. Thus no “inventiveness” could possibly happen no matter what was invented, because all points were reachable within the infinite search space -- it is simply sifting of an ever greater and semi-infinite range of possibilities. See next.
I'm sorry I haven't had time to explicitly address this argument. I'll deal with it briefly. My quick reading of the paper you reference (sorry, no time to go look it up right now) suggests that the flaw in your argument is that it starts with a system in which all possible variants have already been constructed or, to put it another way, a possibility space that is completely unconstrained. Certainly, if all possibilities are in the same hypervolume, then certainly any invention can be generated. But my point is that nature is not like that. In a Darwinian world, biological features must be generated in a step-by-step mechanism from pre-existing biological features, and those features are each produced by a complex developmental process, and are acted upon by the requirement that the results of this developmental process must be able to survive in their environment. These limitations put some severe constraints upon what can be generated from any given starting point. Specifically, I argue that the requirement for viability is what chops the possibility space up into discrete sections or hypervolumes with distinct edges that cannot be crossed by the Darwinian mechanism. As yet, I've not seen any counterexamples that utilize realistic constraints upon the hyperspace of possibilities to show how one moves from one hypervolume to another via a trial and error mechanism. Your example fails because it does not properly model the functional constraints that lead to the quantiziation or discretization of possibility space.
ACSC,
Thanks for the insightful comments. You are 99% right, and you seem to understand my ideas better than just about any of my critics. I really appreciate that. You are wondering whether we can really say that any two artifacts/organisms are in different hypervolumes. While some cases are not very clear-cut, there are definitely cases where we can point to both GA's and biological changes that would qualify as inventive. For GA's, consider again Dawkin's biomorph example. The Blue biomorph is clearly in a different hypervolume and is inaccessable from the hypervolume of the initial program. There are clearly two hypervolumes of possibilities here. For biological examples, Dawkins again leads the way with his insight that humans cannot just sprout angel wings from their backs--to do so would require re-engineering the developmental process such that angel wings are possible (in other words, such that this new solution is included in the hyperspace of possibilities). Furthermore, I can see examples where precisely this sort of change has occurred. The origin of developmental programs is precisely this sort of change, where the constraints that map out a hypervolume of possibility come into existence. I map out other examples in my paper. Perhaps you can explain why you feel that my examples can be included in the same hypervolume. How, exactly, is a single-celled organism in the same hypervolume as a metazoan? Surely there are possibilities available to the metazoan body plan which are simply not possible for the single-celled organism? (Things like arms and legs strike me as a good example, or circulatory/respiratory systems).
While it's true that hypervolumes may not always be clear-cut (I can think of examples where I can't clearly delineate between hypervolumes), I think we can point to a few dramatic examples in the history of life. I'll come back to this when I reply to Rex.
Thanks,
John
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 08. March 2003 17:09
Dear John,
You seem to misunderstand my reason for asking for a definition. If you have read my response more carefully you should have noticed that I stated that according to the common usage of the term and according to (my reasonable interpretation) your own usage of the term I believe that I and others have shown that your arguments about hypervolume and GA's has been disproven.
But in order to avoid the possibility that your definition is at odds with my interpretation and the more common usage of the term I asked you to provide us with your definition. In fact such a definition not only would help our discussion but also would make your arguments more rigorous.
I encourage John not to create and attack strawmen which would only serve to further undermine a fruitful discussion on this topic.
Your silence to my question, other than a somewhat ironic strawman ad hominem attack, further suggests that my understanding of the term hypervolume may not be at odds with yours after all. You continue to further muddle the waters by not only refusing such a definition but also to formulate another strawman:
quote:
Basically, all I'm asking is that you show that the algorithm output a solution which was not accessible from the initial program. Show how the program reached a solution that was not part of the original possibility space, that is, there was no way for the program to get to the solution it eventually found.
Of course such a question is not only a non-sequitur but also irrelevant to John's original claims. Of course there is a way for the program to find the solution but the solution belongs to a different hypervolume than the starting conditions. I suggest that John spends some effort on his definition of hypervolume as used in his paper and address my rebuttals of what I understand his usage to be or admit that he was really not arguing for anything more than that GA's cannot find solutions they cannot find without really addressing the issues of inventiveness. Focusing on the biomorph program is not going to be helpful either since the biomorph program did not increase its hypervolume so it is not surprising that it did not find any proposed solutions by John.
Let me once again encourage John to provide for a more rigorous definition of his claims by enlightening us on his usage of the term hypervolume which now seems to mean that "GA's cannot reach solutions they cannot reach" without much further rigorous derivations of these 'concepts'.
I am looking forward to some rigorousness in the terms used by John that may help us understand if my interpretations of his arguments were incorrect.
Nevertheless it may be interesting to note that other authors seem to have made very similar observations to John but rather than rejecting GA's abilities without any rigorous foundation, they have actually shown how such GA's seem to be able to be truely creative and inventive.
Let me also add once again a reference to John's original paper quote:
This observation suggests that we may consider any genetic algorithm to be operating within a certain n-dimensional hypervolume, and certain fixed parameters completely determine that hypervolume ahead of time.
As I have pointed out however there are GA's where the hypervolume is NOT fixed ahead of time which means that not only the parameters are allowed to evolve but also the design space. One could argue that such GA's will still find solutions in their 'hypervolume' but since this hypervolume has now become unconstrained, this basically means that like the human inventor, the GA found an innovative solution by moving outside its original hypervolume of parameters.
Additional hints to how John interprets hypervolume and which would make my observations and conclusions apply to his usage of the term can be found in the above mentioned paper:
quote:
This hypervolume is fixed by certain non-varying parameters (In Dawkin’s Biomorph example, the number of genes and the rules regarding how the integer values of each gene are interpreted) that an intelligent agent must set and which are not allowed to vary.
Number of genes seems to determine the hypervolume, hence increasing the number of genes would lead to an increase in hypervolume. But what if the number of genes itself is allowed to vary? In fact I have shown several examples of just this. All in all I have to conclude that John's orginal usage of the term hypervolume suggest one very similar to my interpretation. Nevertheless I will allow John to more rigorously define his usage of the term even if this means a definition which is at odds with his original usage of the term. All I am asking for is some rigorousness.
[ 08. March 2003, 17:38: Message edited by: Frances ]
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 08. March 2003 20:17
John wrote:
quote:
Basically, all I'm asking is that you show that the algorithm output a solution which was not accessible from the initial program.
There are no examples of a program producing output that it can't produce, because by definition if it produces the output then it could have produced the output. What are you trying to ask here--it doesn't seem like a sensible question!
quote:
I argue that the requirement for viability is what chops the possibility space up into discrete sections or hypervolumes with distinct edges that cannot be crossed by the Darwinian mechanism.
But you provide no examples of distinct edges--and you seem to regard our examples showing that there is no distinct edge where one might expect one as not even understanding your point.
You do provide examples of vastly distant and complicated systems with an incredibly complicated mess in between them, and speculate that there is no path between these two, despite having only a modest idea of what the systems even are and how they work. (E.g. with multicellularity.) Far from being a distinct edge, this is a hopelessly fuzzy mess.
You claim that there are distinct edges--show some!
quote:
While it's true that hypervolumes may not always be clear-cut (I can think of examples where I can't clearly delineate between hypervolumes), I think we can point to a few dramatic examples in the history of life. I'll come back to this when I reply to Rex.
I await the response. In the meantime, let me mention a few things that I will find utterly unimpressive. In any historical process, you can find distinct extant groups even when there is a known common source. For example, you might look at political party affiliation and note that everyone is either a Democrat, or a Republican, with very little in between. You might therefore conclude that there are exactly two political hypervolumes. Except, oddly enough, both were in many ways intellectual heirs to Jefferson's original Democratic-Republican party (and neither really resembles the original now). Luckily we have history books to tell us these things.
If you look now and say, gee, there are only Prokaryotes, Eukaryotes, and Archaea, and that therefore the three must be in different hypervolumes now and could not have been reached by evolution from a common ancestor, you run the risk of making a Democratic-Republican style mistake.
We can distinguish the somewhat-shared history of D&R from, say, the Communist party by reading history and knowing that we have a fairly complete account.
As such, any "dramatic" examples of different hypervolumes in the history of life will be much more dramatic if they are recent enough so there is reasonable expectation that the history of a divergence would be preserved. Alternatively, the examples will be dramatic if it can be shown in principle that there are no paths in between.
Nelson and I have argued about polarity genes, and it's largely irrelevant, so I am dropping it since I'd just start making points again that I made three posts ago. The posts are too spread out to keep the context in mind. If you are willing to pay my funding for the research if it works out the way I expect, I'll go to the lab next door and demonstrate a 50% lethality phenotype from a defect in developmental gene, if it would help.
We've also argued about what T3SS shows regarding the flagellum. I'm not going to pursue this either because we're going in circles here, too. It's in danger of falling into the "one intermediate doubles your problems because now you have two gaps" fallacy, and theory and observation are being conflated.
The only thing I will say is: Nelson, please demonstrate why, if an export system was a functional intermediate on the way to the evolution of the flagellum, that we would expect to see that export system around now. Or, please demonstrate why an export system with similarity to the flagellum could not function or could not be adaptive.
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 08. March 2003 23:07
The point about telling a story about how a bacteria got a flagella, has to take into account the complex interactions and the irreducible nature of the bacterial flagellum itself. Irreducible complexity is a pattern that is not a fabrication.
Your own post answers your question about the bacterial flagellum not being an example of just any ole unlikely outcome. RBH likens this to the probability of getting any ole hand in a card game.
The improbability of getting any hand of poker is an example of just complexity. When determining whether something is both complex and specified one must be able to detach that particular pattern from the event.
RBH's and your own example only demonstrates that a given poker hand is complex (low probability). But because natural selection's job is to break down vast improbabilities, it doesn't really warrant a design inference. So although a poker dealer has just handed me a set of cards that shows high complexity (low probability) I am pretty confident in attributing it to chance, because there is no specificity. Now, suppose if I told the dealer that I wanted a specific hand. Then the hand would be complex and specified, since I independently identified it by asking for it.
As far as someone just having those charactersistics, that is not a chance event, the improbability is further broken down by the parents characteristics. That is why if your wife mates with a person of a different race, (say you are an American and you get a Chinese baby) you suspect design, because it is way too improbable that you would get a distinctly chinese looking baby when neither of you are chinese (design: your wife fooled around on you). Or you would probably look in your ancestral line to see if you have any asian ancestor that would have caused the characterstics to show up.
Or say that I told you exactly how your baby would look like before it was born. You would either think that I switched babies on you or that I am some kind of genious.
When probabilities descend and they are attached to specifications, we usually attribute such things to intelligent design.
I really don't see how showing the possibilty of an intermediate step results in a more likely pathway. As Dembski stated in his reply to Miller, it is possible that pure chance alone can build the flagellum. But even evolutionary biologists regard this as unlikely and therefore reject it. It is not what is logically posssible but what is empirically probable. We can now consider the possibility space as smaller.
But theres more.
As far as exploring the space of all possible configurations of a bi-directional rotary motor, that any configuration that can get us that function would likely be irreducibly complex itself. It is the property of irreducible complexity itself which is pointing to design and not necessarily the specific components of the bacterial flagellum. In fact, I could build a flagellum myself that does not have the specific components, but building one would likley also exhibit specified complexity. I could also show you different machines that exhibit irreducible complexity, the eukaryotic cilium is not made up of flagellin, but it's still irreducibly complex.
This is why an experiment is suggested by Mike Behe that would start with an export motor, and a pilus, and see if that would get us , not exactly the flagellum (in fact if we do get the exact bacterial flagellum that we see in extant bacteria you can go ahead and throw Darwinism in the garbage, there is nothing random about that event), but something just as irreducibly complex that would confer motility. Getting something irreducibly simple or reducibly complex won't get us anywhere.
By showing that a system exhibits the special case of specificity, irreducible complexity, we have narrowed down the space of possibilities.
quote:
What justifies us in attributing specified complexity to the bacterial flagellum? The bacterial flagellum is irreducibly complex, meaning that all its components are indispensable for its function as a motility structure. What's more, it is minimally complex, meaning that any structure performing the bacterial flagellum's function as a bidirectional motor-driven propeller cannot make do without certain basic components.
Consider David Ussery's and Thornhill's paper I mentioned before. Already we see that it eliminates several possibilities to get at irreducible complexity. We see a narrowing down of that space of possibilities, making the end result less and less likely.
Dembski continues:
quote:
In assessing whether the bacterial flagellum exemplifies specified complexity, the design theorist is tacitly following Earman's guidelines for making an eliminative induction work. Thus, the design theorist orders the space of hypotheses that naturalistically account for the bacterial flagellum into those that look to direct Darwinian pathways and those that look to indirect Darwinian pathways (cf. Earman's requirement for an ordering or topology of the space of possible hypotheses). The design theorist also limits the induction to a local induction, focusing on relevant hypotheses rather than all logically possible hypotheses. The reference class of relevant hypotheses are those that flow out of Darwin's theory. Of these, direct Darwinian pathways can be precluded on account of the flagellum's irreducible and minimal complexity, which entails the minuscule probabilities required for specified complexity. As for indirect Darwinian pathways, the causal adequacy of intelligence to produce such complex systems (which is simply a fact of engineering) as well as the total absence of causally specific proposals for how they might work in practice eliminates them. In eliminating indirect Darwinian pathways, design theorists are therefore not merely eliminating what thus far hasn't worked (coevolution and co-optations) but also appealing to causal powers (designing intelligences) that are known to work.
Now in my opinion, the possibility that the bacterial flagellum arose from a primitive export machine has been effectively eliminated due to the amount of pure chance events invoked to get the job done. It is relevant, in telling these stories, to bring up the interactivity of each particular component of the flagellum.
quote:
Science must form its conclusions on the basis of available evidence, not on the possibility or promise of future evidence. This means that eliminative inductions need to be local inductions, based on detailed testable models and hypotheses that are currently available.
[ 08. March 2003, 23:39: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 08. March 2003 23:18
Rex:
The only thing I will say is: Nelson, please demonstrate why, if an export system was a functional intermediate on the way to the evolution of the flagellum, that we would expect to see that export system around now.
Nelson:
This is indicative of the trial and error process of a co-option explanation. Plasticity is a prediction of co-option. Furthermore, there is no reason to think we shouldn't have evidence of intermediates since selection would have conserved the initial non-flagellar function, it was selectively advantageous to have an export machine in the first place, before in some organism, addition of another mutli-part machine, by pure chance, changed the function.
Rex:
Or, please demonstrate why an export system with similarity to the flagellum could not function or could not be adaptive.
Nelson:
Thats the problem. Again, the likely explantion for the similarity is that it evolved from the flagellum and current papers are agreeing with this. Secondly, that an export system would function and would be adaptive is the problem , as selection is a conservative process and a pure chance event would have to break free the function of the export machine from the claws of natural selection, not just one, but multiple events.
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 08. March 2003 23:57
Well, this thread has splintered into so many directions I think I will only comment on what I think is the most important point that has emerged: I had John Bracht's definition wrong. I thought he was identifying "inventiveness" based on features that could be identified independent of knowledge about their evolvability. Thus we would have an actual legitimate argument that can be assessed:
1) Inventive solutions, identified by features A, B, and C, have occurred in the history of life
2) Evolution can't produce inventive solutions because of Y, Z
3) ID can produce inventive solutions
4) Therefore ID occurred somewhere.
But in fact, all along he was arguing that a program can't reach what a program can't reach:
quote:
Basically, all I'm asking is that you show that the algorithm output a solution which was not accessible from the initial program. Show how the program reached a solution that was not part of the original possibility space, that is, there was no way for the program to get to the solution it eventually found.
I agree with him. By definition, a program can't reach what it can't reach.
What this does, however, is make his argument circular, just like Dembski's specified complexity argument, many versions of Behe's argument, kinds/baramins, etc.
In other words:
1) Someone claims that evolution cannot produce systems matching criteria X (e.g. SC/IC/inventiveness/information).
2) If evolution is shown to reasonably produce what seems to be X, then it emerges that the *actual* definition of X includes "evolvability" as *part of the definition*, and therefore whatever evolution can do, it isn't X. This of course begs the question of whether X under the "evolvability" definition ever actually happened in the history of life.
All this proves, as a talk.origins poster long ago wrote, is that a circular argument is irreducibly complex.
How many times shall we go on this same merry-go-round, I wonder?
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 09. March 2003 00:07
Nelson:
The point about telling a story about how a bacteria got a flagella, has to take into account the complex interactions and the irreducible nature of the bacterial flagellum itself. Irreducible complexity is a pattern that is not a fabrication.
And how do we determine whether or not the pattern is not a fabrication? I would say that since the probability of the system would depend on our knowledge we may have a problem. In fact as Sobel has argued, specification is trivial for almost any event, making many chance events appear to be designed.
Nelson's example about the poker hand seems irrelevant since in the case of the flagellum the specification is NOT done after the fact. But using Sobel's approach one can generate specification which includes all possible hands in poker hence the issue of specification seems to be trivial.
The case of the flagellum is more like a poker hand which has been dealt which now needs to be specified. For many hands I am sure that one can find some card game in which such a hand was dealt. From Sobel
quote:
From this second illustration can be gathered that Dembski's theory enables a moderately imaginative person, with a list of possible delimitations of an event, easily eliminate relevant chance-hypotheses for the event; if they all make more probable that not its non-occurrence, and avoiding 'false negatives' concerning relevant chance- hypotheses for this event is somewhat (it need not be very) important to him. From the two illustrations, one may gather that by the lights of Dembski’s book, we are entitled, and will always be entitled to conclude, that not much happens by chance.
Thus when Nelson states "When probabilities descend and they are attached to specifications, we usually attribute such things to intelligent design." he has to take into consideration the likely false positives. Dembski seems to be struggling with this concept, on the one hand he seems to accept false positives but on the other hand he has to reject any such occurrence for the design inference to have any practical value.
Thus when Nelson states that "I really don't see how showing the possibilty of an intermediate step results in a more likely pathway. As Dembski stated in his reply to Miller, it is possible that pure chance alone can build the flagellum. But even evolutionary biologists regard this as unlikely and therefore reject it. It is not what is logically posssible but what is empirically probable. We can now consider the possibility space as smaller."
And yet showing intermediates does reduce the complexity of the whole system. If Dembski is arguing against a pure chance flagellum construction then I wonder why he spent so much effort on this strawman when calculating the chance probabilities for the flagellum?
So far the mere possibility of indirect and even direct Darwinian routes suggests that ICness is not a helpful concept in determining likelihood of design. Dembski may want specific scenarios but specific scenarios do not make an ID inference more or less likely or do they? Nelson may object to the stories so far but what Nelson cannot do is conclude that there is evidence of design in the flagellum since that depends on showing that no indirect pathways to IC systems exist.
At the moment the best we may be able to suggest is: "We don't know". I doubt that this would be an acceptable design inference for most people :-)
Or as Nelson states it so clearly:
quote:
Science must form its conclusions on the basis of available evidence, not on the possibility or promise of future evidence. This means that eliminative inductions need to be local inductions, based on detailed testable models and hypotheses that are currently available.
Science thus must form its conclusion s not on the possibility or promise of future evidence of intelligent design. This means that eliminative inductions need to be local inductions, based on detailed testable models and hypotheses. But Dembski also suggests that such models are not likely forthcoming from an intelligent design perspective making the concept of ID not very attractive as an example of what Dembski believes science should be doing.
So when Nelson states that
"Now in my opinion, the possibility that the bacterial flagellum arose from a primitive export machine has been effectively eliminated due to the amount of pure chance events invoked to get the job done."
We seem to be able to identify the following problems 1) Pure chance is not invoked by such mechanisms 2) Since we do not fully know the full pathways for likely events we cannot ascribe any probabilities.
I may agree with Dembski that detailed evolutionary pathways may be lacking but ID seems to suffer from similar and if I understand Dembski correctly unsurmountable problems.
[ 09. March 2003, 00:34: Message edited by: Frances ]
IP: Logged
|
|
Cre8ionist
Member
Member # 140
|
posted 09. March 2003 09:25
Hi, not going to jump into the mix more than once. I took the time to read most of the posts (a huge feat) and many of the links and wanted to make a couple of comments and ask a couple of questions.
First, I don't think John is saying that a program can't accomplish what it can't accomplish, but that a program can't accomplish what Darwinian Evolution is said to have accomplished. A not so subtle distinction.
Further, I don't think the argument is irreducible, on the contrary, it's quite reducible, so much so that we could put it in much simpler terms. It seems that it's really the old quality/quantity argument in another form.
Correct me if I'm wrong but isn't John basically saying that the hypervolume, while able to increase in quantity, does not increase in quality? In other words, it increases horizontally but not upward/vertically? And isn't this basically the same argument that Spetner made about
information a few years ago?
Finally, isn't this testable? If not now, soon? Wouldn't a program (to seriously mimic Darwinian Evolution), need to be able to do things like write new functions through trial and error? So at least in theory, be able to turn itself from a black and white program to a color program as John has alluded to? Otherwise it is only doing what it was programmed to do and nothing really like the amoeba to man theory of DE. I mean if Dawkins' program always puts out the same "kind" of critter and flies always remain flies we've got a serious challenge to overcome if we want to say that DE is the originator of new critters. Anyway, great thread Frances...................................Cre8
IP: Logged
|
|
John Bracht
Member
Member # 5
|
posted 09. March 2003 12:53
Hi all,
No time to write much now, but this thread needs clarification, fast.
Rex, Yersinia are taking my comments entirely out of context. If they had read more carefully, they would have seen this sentence following closely to the one they quote:
quote:
Show how the parameters of the system were adjusted purely by the GA itself to allow the novel solution to be found (a solution that was not originally reachable, in any way, with the initial parameters).
Precisely as Cre8 said, show how the program re-writes itself such that a new hypervolume of possibility is created. Show that the program altered itself (in other words, alters its own encoding) such as to allow for new possibilities to be reached that couldn't be reached when the program began.
Francis seems to have given up on this, having spent thousands of lines of text arguing that his examples have shown just this. Now, he claims to have been using a different definition of hypervolume. Well, all I can say is that this is a great stawman, to erect your own definition to a term in an argument you're trying to defeat, then knock it down. Of course, if you define hypervolume in a different way (and a rather ad-hoc way, it appears) you can show that programs increase it. But if you use it as I've been using it, consistently, all along (as the space of possibilities that a program can explore, given its encoding), then I think it's pretty clear that GA's and trial and error processes can't increase it. Inventiveness would alter the encoding itself such that new possibilities are reachable. That's all you need to show, Francis--I've made it as clear as I know how. The definitions really aren't that hard or confusing--I think you're making things way more confusing than they really are. If you're still wanting to dialogue, I would like to see how the Guo et al program does this sort of inventive change.
John
[ 09. March 2003, 12:55: Message edited by: John Bracht ]
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 09. March 2003 15:08
Seems to me that we've got it exactly right.
quote:
Rex, Yersinia are taking my comments entirely out of context. If they had read more carefully, they would have seen this sentence following closely to the one they quote:
quote:
Show how the parameters of the system were adjusted purely by the GA itself to allow the novel solution to be found (a solution that was not originally reachable, in any way, with the initial parameters).
Precisely as Cre8 said, show how the program re-writes itself such that a new hypervolume of possibility is created. Show that the program altered itself (in other words, alters its own encoding) such as to allow for new possibilities to be reached that couldn't be reached when the program began.
This is a fool's errand. You are defining the "hypervolume" as (paraphrase) "that which was reachable with the initial parameters". Obviously with such a definition, a GA can't escape it's hypervolume. What most people would consider "rewriting" -- i.e., increasing the functionality encoded in the genome via the evolution of new genes, organs, etc. -- is, under John's definition, all part of the same "hypervolume", because it is in fact reachable.
If it turned out that evolution was able to turn a single cell into a redwood, this would still be part of the same hypervolume, because John defined it that way!!.
This is essentially a replay of Dembski's specified complexity argument at this point, which boiled down to "Wildly improbable/impossible things are wildly improbable/impossible". True, but this does nothing to establish that the evolution of novel complex systems is improbable.
[ 09. March 2003, 15:23: Message edited by: yersinia ]
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 09. March 2003 15:37
Creationist:
First, I don't think John is saying that a program can't accomplish what it can't accomplish, but that a program can't accomplish what Darwinian Evolution is said to have accomplished. A not so subtle distinction.
That is however not what John seems to be saying. If that is what he wants to say then I suggest that he clarifies his usage of terms. So far John's argument seems to be not much different from a circular argument.
quote:
Show how the program reached a solution that was not part of the original possibility space, that is, there was no way for the program to get to the solution it eventually found.
I have no idea how you define horizontal and vertical in a multi-dimensional concept nor how you define quality versus quantity. Certainly any link with Spetner is totally lost on me. As I understand Spetner's 'argument' mutation cannot increase information (despite evidence to the contrary). How does this come into play here?
I believe that not only is the concept testable but it has been tested and shown that according to the common intepretations of creativity and hypervolume, GA's indeed not only can increase their hypervolume but also be 'creative'.
From Gero
quote:
Innovative designing, in computational terms, can be defined as that designing activity that occurs when the constraints on the available ranges of the values for the variables are relaxed so that unexpected values become possible,
quote:
Creative designing in computational terms, can be defined as the designing activity that occurs when one or more new variables is introduced into the design. Processes that carry out this introduction are called “creative designing processes”. Such processes do not guarantee that the artifact is judged to be creative, rather these processes have the potential to aid in the design of creative artifacts. Thus, creative designing, by introducing new variables, has the capacity to produce novel designs and as a result extends or moves the state space of potential designs.
A program which not only found the optimal parameters but also the optimal topology:
quote:
This paper has demonstrated that genetic programming can be used to automatically create both the parameter values tuning and the topology for controllers for illustrative problems involving a two-lag plant and a three-lag plant.
Gero comments on what needs to happen for non-routine design
quote:
Non-routine designing maps onto creative designing. In routine designing all the variables which specify designs are given in advance. This means that the space of possible designs is known a priori, each point in this space can be constructed and evaluated directly. What needs to be done is to search this space in order to locate an appropriate or most appropriate design. The result here is the “best” design from this space. In nonroutine designing the result is the “best” space of possible designs as well as the “best” design from this space. Processes which modify the design space of the search problem are called exploratory processes.
quote:
One of the well-established notions related to creative designing processes is that an important means of characterising them is to determine whether they have the capacity to expand the state space of possible designs - exploration (Gero, 1994).
So in other words non-routine design requires not only that the optimal parameters are found but also that state space of possible designs can be expanded. Note that these are requirements but not necessarily sufficient in that GA's which can expand their state space need not reach creative solutions. Nevertheless it does show, unless John wants to provide us with _his_ definition of the term hypervolume (in a non tautological manner), that one cannot beforehand rule out creative design from GA's based on the argument of hypervolume. One can rule out certain classes of GA's being able to expand their hypervolume but not in general.
Thus, whether or not evidence of innovative/creative solutions exist in biology does not help us eliminate in one grand swoop (to use Dembski's terminology) GA's as their 'creators'. While it may have seem attractive to be able to dismiss, similarly to for instance IC or CSI, large sets of pathways, it seems that like IC and CSI, the concept of hypervolume suffers from vary similar shortcomings.
I find it frustrating to have to argue John's strawmen arguments. Now he seems to argue that I claim to be using a different definition of hypervolume. I understand that the pressure of impending final exams may limit the time John can spend on this topic but as with final exams I believe that the qulaity of the answer depends critically on reading the text carefully.
What I did say is that I used the common definition of hypervolume, a definition which seemed to match with John's usage. But in order to avoid John arguing that his usage of the term hypervolume is different I encouraged John to provide us with a more rigorous definition and approach to his argument. Instead John seems to have chosen to not address the argument but to go after the person delivering the bad news. That's not a dialogue in any form or manner. Until John defines his terms in a more rigorous manner, I predict that our 'discussions' will not be very fruitful in exploring his arguments.
Thus until John provides me with his definition of hypervolume and shows that it is different from my interpretation and the common usage of the term, I will point out that according to the common usage of the term GA's not only can expand their hypervolume but in fact they have been shown to actually do so leading to creative, non-routine design solutions.
As Yersinia has shown us so pointedly, John's latest 'definition' of hypervolume seems to be more like a self fulfilling prophecy rather than a rigorous (mathematically speaking) definition.
IP: Logged
|
|
|
Printer-friendly view of this topic
