|
Author
|
Topic: Dembski on functional subsystems in "Uncommon Dissent"
|
Nel
Member
Member # 614
|
posted 25. July 2003 22:17
PvM,
You had recently stated:
quote:
Thus when Nelson states "Complex systems like flagella are harder, because of what Dembski calls the "sea of non-functionality"." he lacks any supporting data about this 'sea of non-functionality' in fact this sea seems to be more like a dessert to me
You even put a little smiley face at the end. After three replies, you still have not substantiated this claim. If you can't why did you say it?
[ 25. July 2003, 22:18: Message edited by: Nelson-Alonso ]
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 25. July 2003 23:24
quote:
Nic writes:
quote:
--------------------------------------------------
Nelson, Dembski says that subsystems "always" exist! What more do you want?
--------------------------------------------------
Of course subsystems always exist. As I already discussed, thats not what matters. What matters is how much of a gap there is between subsystems, such that a series of parts is unselectable. If a system can be decomposed to simpler functions, it can likely evolve. If only one (as in the case of flagella), then I'm afraid the sea of nonfunctionality needs to be dealt with.
Sigh. Once again Nelson is trying to establish something about the flagellum via repeated assertion.
Let's see how many subsystems (sub-subsystems) with different functions I can identify, off the top of my head:
Type III secretion system
- FliI (hexameric ATPase, similarly widespread)
- FliF (inner membrane pore -- there are a large variety of transmembrane systems, of varying complexity ranging from passive pores all of the way to ATP-powered protein secretion systems)
- Peptidoglycan pore, type II SS-secreted -- such things are useful any time an export system begins to export proteins big enough that diffusion through the cell wall is slowed down
- Outer membrane secretin & lipoprotein chaperone (widespread homologs in many secretion systems), type II SS-secreted
- Detailed function and structure of other T3SS components still almost completely unknown, so declarations about their nonfunctionality in potential independent subsystems is wild bluster
Motor proteins- proton-conducting ion channel: MotAB homologs (ExbBD, TolQR)
- ExbBD and TolQR each perform work on a third protein; motAB performs work on FliG. Perhaps FliG originated as a fragment of a substrate protein that happened to bind to FliF. The proto-MotAB would then automatically associate with this proto-FliG
Axial protein family- rod, hook, adaptor, filament, and cap proteins probably all descended from a common ancestor; the filament of type III pili today thus far has only one clear component
- muramidase domain of FlgJ has widespread homologs
Switch complex- FliN and FliM are homologs of each other, the receptor domain on FliM is homologous to a free-floating, single-domain chemotaxis protein, CheC. The proto-FliM might have started out as a membrane receptor.
- Switching is in any case dispensable for dispersal, and reversability is similarly dispensable, many bacteria get by with stop-start flagella
Chemotaxis- Chemotaxis system is part of general cellular environmental response system, and is independently attached to many different forms of motility
- Some prokaryotes appear to lack the standard chemotaxis system entirely
- Would be unnecessary anyway for functions such as dispersal
- The core of the chemotaxis system is the two-component signal transduction system which is universal and found in numerous non-motility systems
Well, that about all of the typically-listed components, actually. The only really tough components to identify functional subsystems for are (*some*) of the type III export apparatus components, but then their current function is pretty much unknown. And besides, the challenge "explain the flagellum" does not require one to trace the ultimate origin of every subsystem all the way back to the origin of life -- such requirements are just another instance of moving the goalposts once one challenge has been met.
Like the EQU function, which can evolve with 5-7 simpler subsystems available, for the flagellum I have identified 5 major subsystems and several additional sub-subsystems. A "sea of non-functionality" even at this early stage of investigation this does not look likely, despite Nelson's continual assertions.
(BTW, you guys really should learn how to spell "desert"...a "dessert" is something rather different. :-) Hmmm, I'm feeling hungry... )
[ 25. July 2003, 23:31: Message edited by: yersinia ]
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 25. July 2003 23:33
A couple of days ago Nelson Alonso wrote to me quote:
Show me a case where only two simpler functions were rewarded and yet they still obtained EQU.
And responding to yersinia, Nelson Alonso wrote quote:
But if you can only decompose it to very few simpler functions, then you aren't really negating the fact that you are ushering in too many chance events. In fact, as RBH notes, EQU didn't evolve when you took away too many intermediates. Thats the problem that IC points to.
What I actually said was that the Lenski, et al., control conditions show that if one takes away all intermediate functions then EQU doesn't evolve. Take away one or two intermediates and lineages that perform the input-output mapping corresponding to EQU still evolve. "Too many" is Nelson's gloss of "all." He keeps trying to invoke some sort of 'quantity of intermediates' variable, apparently with a threshold below which there are said to be too few to enable evolution. I strongly suspect that "too few" will always be one less than the minimum that current research shows does enable evolution of complex functionality in a particular system.
As I mentioned somewhere earlier, I'm doing familiarization runs with both versions 1.3.0 and 1.6.0 of Avida, replicating several of the Lenski, et al., conditions and generally getting familiar with the program, its control files, and its output files. This familiarization process is necessary because the documentation on the Web is for version 1.0.0, and some things have changed.
While I haven't done a two-intermediate run in the course of program testing and familiarization, I have done several very long (500,000 cycles) runs rewarding only three 'intermediate' functions. One of them evolved to perform the input-output mapping corresponding to EQU in less than 500,000 cycles. (Recall that Lenski, et al., used runs of 100,000 cycles maximum.) The test run that evolved a lineage that performed the EQU input-output mapping had a fitness function that rewarded only NOT, OR_N, and AND_N (and EQU, of course) with the same weightings as in Lenski, et al. OR_N is a 2-nand operation and AND_N is a 3-nand operation. The other two-input mappings (corresponding to the 1-nand NAND itself, 2-nand AND, 3-nand OR, 4-nand NOR, and 5-nand XOR) were not rewarded; fitness bonuses for them were commented out of the appropriate program control file.
As I said, I'm still familiarizing myself with the control files and the various output files that can be dumped to disk and with analyzing those data, but I'm sure the result from the run that evolved to perform EQU is valid. The TASKS.DAT output file, which records how many genotypes perform each task every 10 cycles (as I had it set for these runs) shows it, and EQU's appearance is consistent with patterns in the fitness data file for the dominant creature, also recorded every 10 cycles during the run.
Any other questions? Be aware that I don't promise to do the runs necessary to answer any or all questions, since I am not particularly interested in spending several hundred hours or more testing and refuting every damfool offhand claim that anyone might make about the Lenski, et al., study. Depending on the question, I might do the necessary runs or I might not. Most likely I'll probably say something like 'Download the program and spend your own CPU cycles! And if I see a report of results that I think is dubious, I'll replicate it to see if it holds up.'
Finally, Nelson Alonso wrote quote:
RBH: quote:
By the way, when does Nelson envision that "ID theory" will present a plausible and perhaps even testable historical narrative of the origin of the flagellum?
When will you reply to the many responses you've been given to this question?
When I see one that's responsive to the question.
RBH
Added in edit (after reading yersinia's posting after I posted this): That a 3-intermediate lineage evolved to perform the EQU mapping means that yersinia's "5-7" reduces to 3 now.
[ 25. July 2003, 23:40: Message edited by: RBH ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 26. July 2003 00:13
Nic writes:
quote:
Type III secretion system
Yes this is one possible alternative function. Unfortunately, it's been shown and there are good reasons to think, that T3SS evolved from flagella. So the pointing them is moot.
Nic writes:
quote:
proton-conducting ion channel: MotAB homologs
There is no such thing as MotAB homologs. MotA is homologous only to ExbB.
Nic writes:
quote:
(ExbBD, TolQR)
ExbBD and TolQR each perform work on a third protein;
No. The similarity is just between MotA and ExbB, and does not involve FliG, which seems not to have a counterpart in the Exb system. TonB might be involved in forming part of the channel, or causing a structural change that opens the channel. MotAB/fliG seems to have no non-flagellar function.
Nic writes:
quote:
Axial protein family
rod, hook, adaptor, filament, and cap proteins probably all descended from a common ancestor;
Thats just an assertion.
With regard to the switch complex,
Nic writes:
quote:
FliN and FliM are homologs of each other, the receptor domain on FliM is homologous to a free-floating, single-domain chemotaxis protein, CheC.
The switch complex is actually made up of FliG, FliM, and FliN and are required in the bacterial motor. What is the % sequence similarity between FliN and FliM?
Nic writes:
quote:
Switching is in any case dispensable for dispersal, and reversability is similarly dispensable, many bacteria get by with stop-start flagella
I'd like to see a reference on this one. I think flagella are required for dispersal.
The chemotaxis system was never scored as part of flagella AFAIK. So it is irrelevant. That still leaves 1 alternative function, type III secretory system.
Also, you are comparing the number of intermediates required for EQU with 3-dimensional bacterial flagella. Of course the comparison is ridiculous. But it illustrates my point nonetheless. However, saying 5 are required for EQU is not to say that 5 would be enough for flagella.
[ 26. July 2003, 00:21: Message edited by: Nelson-Alonso ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 26. July 2003 00:16
I'll respond briefly to RBH by making the same point, then I'm off to bed.
RBH writes:
quote:
While I haven't done a two-intermediate run in the course of program testing and familiarization
Actually you would have to do a 1 intermediate run if you are comparing the flagellum to EQU literally. However, as I already said, also, you are comparing the number of intermediates required for EQU with 3-dimensional bacterial flagella. Of course the comparison is ridiculous. But it illustrates my point nonetheless. However, saying 5 are required for EQU is not to say that 5 would be enough for flagella.
quote:
That a 3-intermediate lineage evolved to perform the EQU mapping means that yersinia's "5-7" reduces to 3 now.
How many instructions did EQU have in this run? Can you reference it? Which 3 simpler functions did it use as intermediates?
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 26. July 2003 01:45
Nelson wrote quote:
Actually you would have to do a 1 intermediate run if you are comparing the flagellum to EQU literally. However, as I already said, also, you are comparing the number of intermediates required for EQU with 3-dimensional bacterial flagella. Of course the comparison is ridiculous. But it illustrates my point nonetheless. However, saying 5 are required for EQU is not to say that 5 would be enough for flagella.
Once again, I have not made and do not make that direct comparison! Please stop attributing it to me. As I have said before, my interest (in this context) is in how complex systems and features arise from the local processes invoked by evolutionary theory and in what sorts of population dynamics arise from those local processes. The Avida simulation is a useful test bed to study those questions. As far as I can see, Nelson Alonso is the only one taking (forcing?) the Avida/flagellum comparison literally. And I note that (as I predicted) Nelson Alonso's threshold has dropped from two intermediates for a literal comparison (in his 23. July 2003 21:18 posting in this thread) to just one in his most recent posting.
Nelson asked quote:
How many instructions did EQU have in this run? Can you reference it? Which 3 simpler functions did it use as intermediates?
I don't know how many instructions the lineage that evolved to perform the input-output mapping corresponding to EQU used to do that mapping - I didn't dump the individual lineages to files for that run nor for the others in that series. I ran them for a different purpose (to compare the relative utility and fidelity of several sampling frequencies for discriminating patterns of population changes in averaged data) and so didn't save out the individual lineages. Once again, these are familiarization and exploratory runs in preparation to do some research, not the research itself. However, the averaged data show that the population averaged 136.3 instructions per critter with an average of 61.8 executed instructions per critter around the time when the lineage that performed the EQU mapping appeared. The remainder were 'junk' instructions. There's no way to know from the available data how many of the executed instructions were involved directly in performing the EQU mapping.
The run that evolved to perform the EQU mapping (as I said in the original posting above where I described it) used NOT, a single-input function, and OR_N and AND_N, both two-input functions, as rewarded intermediates. The remainder of the two-input logic functions were not rewarded with reproductive resources; their reward functions were disabled. Genotype length was rendered selectively neutral by awarding SIPs in proportion to length, as in the Lenski, et al., procedure.
RBH
IP: Logged
|
|
|
|
Pim van Meurs
Member
Member # 541
|
posted 26. July 2003 12:20
I stated: Thus when Nelson states "Complex systems like flagella are harder, because of what Dembski calls the "sea of non-functionality"." he lacks any supporting data about this 'sea of non-functionality' in fact this sea seems to be more like a dessert to me
Nelson: You even put a little smiley face at the end. After three replies, you still have not substantiated this claim. If you can't why did you say it?
I put a smiley after it because I was chuckling inside. I did not think that we would hear anything substantial about this sea of non-functionality and the term 'just so story' came to mind. Hence the smiley.
I presume that you have no additional data to support the metaphor used by Dembski either?
Nic has a point "Sigh. Once again Nelson is trying to establish something about the flagellum via repeated assertion."
[ 26. July 2003, 12:21: Message edited by: Pim van Meurs ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 26. July 2003 22:56
RBH writes:
quote:
As far as I can see, Nelson Alonso is the only one taking (forcing?) the Avida/flagellum comparison literally.
Then like I said, and many other IDers have said,Avida is simply irrelevant to IC systems that Behe/Dembski are referring to, and this is precisely what Dembski stated in his OP over at ARN. Now do you get it?
RBH writes:
quote:
And I note that (as I predicted) Nelson Alonso's threshold has dropped from two intermediates for a literal comparison (in his 23. July 2003 21:18 posting in this thread) to just one in his most recent posting.
Actually, when I said 2 rewarded functions, I was counting both T3SS and the fully functioning flagella. However, if you are counting intermediates it is not correct to call flagella an intermediate to flagella, that would non-sensical. So that leaves 1 alternative function.
I had asked:
quote:
How many instructions did EQU have in this run? Can you reference it? Which 3 simpler functions did it use as intermediates?
RBH wrote:
quote:
The run that evolved to perform the EQU mapping (as I said in the original posting above where I described it) used NOT, a single-input function, and OR_N and AND_N,The remainder of the two-input logic functions were not rewarded with reproductive resources;
Is it at all possible for you to e-mail me some of this data? And is it possible to find out how many instructions EQU had? And if the other two-input logic functions arise by chance, after which you rewarded EQU? I find it completely outlandish that NOT and OR_N and AND_N is sufficient for EQU, in which case, the program is even more plastic then even I had thought.
I may have some insight that is relevant to IC with this data. Are you planning to publish it?
[ 26. July 2003, 23:04: Message edited by: Nelson-Alonso ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 26. July 2003 22:58
Nic writes:
quote:
Arrgh! Nelson, *every* study I've read considers the MotAB pair to be homologous to the TolQR pair and ExbBD pair. E.g.:
Cascales E, Lloubes R, Sturgis JN. The TolQ-TolR proteins energize TolA and share homologies with the flagellar motor proteins MotA-MotB. Mol Microbiol. 2001 Nov;42(3):795-807.
It's right there in the friggin' title, what more do you want? Dembski seems to want us to follow the literature, but you seem to wildly disregard it.
Nic, the paper states:
quote:
revealed that the TM domains of TolQ and TolR present structural and functional homologies not only to ExbB and ExbD of the TonB system but also with MotA and MotB of the flagellar motor.
TolQ and TolR have homologies to MotA and MotB and to ExbB and ExbD, that does not necessarily mean that ExbD is homnologous to MotB.
However, if you meant to say that TolQ and TolR are homologous to MotA and MotB, then you should have said that explicitely, what you wrote above was misleading when you included the ExbB system.
[ 26. July 2003, 23:01: Message edited by: Nelson-Alonso ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 26. July 2003 23:03
PvM writes:
quote:
I presume that you have no additional data to support the metaphor used by Dembski either?
If you can provide support for your desert claim, then that would be data against the "sea" metaphor. However, you have yet to substantiate that claim. AFAIK, the export machine is the only "stand alone" function within flagella.
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 26. July 2003 23:33
Nelson writes,
quote:
Nic, the paper states:
quote:
--------------------------------------------------
revealed that the TM domains of TolQ and TolR present structural and functional homologies not only to ExbB and ExbD of the TonB system but also with MotA and MotB of the flagellar motor.
--------------------------------------------------
TolQ and TolR have homologies to MotA and MotB and to ExbB and ExbD, that does not necessarily mean that ExbD is homnologous [sic] to MotB.
Oh, come on, Nelson! If X is homologous to Y, and Y to Z, then X is homologous to Z!
quote:
http://tcdb.ucsd.edu/tcdb/tcfamilybrowse.php?tcname=1.A.30
UCSD Transport Protein Database
The Saier Laboratory Bioinformatics Group
1.A.30 The H+- or Na+-translocating Bacterial Flagellar Motor 1ExbBD Outer Membrane Transport Energizer (Mot-Exb) Superfamily
The Mot-Exb Superfamily consists of two distant families, each with a distinct function. The Mot family energizes bacterial flagellar rotation while the Exb family energizes accumulation of large molecules (i.e. iron-siderophores, vitamin B12, DNA from phage and colicins) from the external medium across the outer Gram-negative bacterial membrane into the periplasm. The pmf (or smf) is the driving force in both cases. MotAB and (PomAB) are homologous to ExbBD and TolQR.
(last bold added)
Nelson continues,
quote:
However, if you meant to say that TolQ and TolR are homologous to MotA and MotB, then you should have said that explicitely, what you wrote above was misleading when you included the ExbB system.
No, it wasn't. I do try to have some vague idea of what I'm talking about...
[ 27. July 2003, 00:39: Message edited by: yersinia ]
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 27. July 2003 03:44
Let me see if I can organize this so it's comprehensible.
1. Regarding comparisons of the Lenski, et al., study with the flagellum: By "literal comparison" I mean attempting to establish a one-to-one correspondence between specific parts of the two systems. The Lenski, et al., study did not purport to provide an isomorphic representation of the flagellum. What it shows is that a system that meets the operational definition of "irreducible complexity" can readily evolve under biologically plausible circumstances. Dembski's OP on ARN does not mention the flagellum. It faults the study for using an irreducibly complex system that can be "decomposed into simpler features exhibiting simpler functions." yersinia has pointed out the glaring contradictions in Dembski's recent statements on that topic and I don't need to add to his remarks. With respect to Dembski's remarks about how (or whether) the Lenski, et al., simulation maps onto biology, I'll say only that it's one heckuva lot better model of biological reality than a mousetrap!
2. As I said, I am doing various series of pilot runs to explore the program, familiarize myself with its control files and output files, and settle some methodological and configuration questions. The particular run I described above was part of a short series looking at a range of sampling frequencies for their fidelity of representation of population dynamics. I used the three simple functions - NOT, OR_N, and AND_N - as rewarded functions along with EQU purely to have a limited set of functions so the population dynamics would (hopefully!) be simpler and I could make some decisions about sampling frequencies for use in formal studies. I was not 'trying' to evolve EQU; I just wanted something - anything - to evolve so I'd have a picture of the representation of changing population dynamics at the various sampling frequencies.
Since I was (and still am, as far as that goes) focused on population dynamics arising from local processes in evolution, I did not save to file detailed information on all the lineages that evolved. Hence it is impossible to ascertain the genotype of the lineage that evolved to perform it. Since the run in which it appeared happened to use a fairly coarse sampling scheme, one can't even tell exactly when it occurred. That it occurred is indisputable; precisely when it occurred and what instructions comprised the critters that did it was impossible to recover from the available files.
3. There seems to be an impression on Nelson's part that more complex features or processes are necessarily evolved by coopting simpler functions whole. Hence his comment that he finds it "completely outlandish that NOT and OR_N and AND_N is sufficient for EQU, ...". But evolution doesn't have to use a whole simpler function in generating a more complex function. The simpler functions provide organized raw materials for more complex functions. In evolving the simpler functions, chunks of code are organized under selective pressure. The input-output mapping corresponding to NOT requires (at least) 6 instructions to perform. The AND_N mapping requires at least 10. OR_N's mapping requires 7 instructions. Substrings within those longer programs are available for cooption; evolution doesn't have to use the whole simpler function. So the question for Nelson's incredulity is whether the evolved programs that perform the NOT, OR_N, and AND_N mappings contain substrings that are cooptable to perform the EQU mapping. The information necessary to evaluate that can be recovered from appropriately configured Avida runs if one is sufficiently interested to do the work.
That this sort of cooption occurs is suggested by the curves describing the tasks performed by the genotypes in the population through time. Not infrequently, as a somewhat more complex function appears and becomes widespread in the population, a simpler function decreases in frequency for a time, sometimes for hundreds of generations, until later evolution recreates it in the new species of critters. For example, in another exploratory run replicating the pattern of the Lenski, et al. fitness function with attenuated rewards, critters that performed the NOR mapping appeared and began to increase rapidly in the population around cycle 220,000, and simultaneously the frequency of critters that performed OR, which had been quite common in the population, decreased by about 25%, only to recover around cycle 280,000. That pattern is explicable if the lineage that evolved to perform the NOR mapping did so by coopting a chunk of the code that it used to do the OR mapping, thereby rendering it unavailable for the OR mapping. Since NOR is more reproductively valuable than OR, that lineage prospered, decreasing the frequency of OR-performing critters in the population. Later, when NOR-performing critters evolved (again) the ability to perform the OR mapping, it recovered in the population.
That association of the emergence of a more complex function with a decrease in the frequency of a simpler function suggests the kind of cooption I've suggested: When part of a simpler function is coopted by a more complex function it may no longer be available to the simpler function: the precursor may in effect be destroyed in the process of cooption. It's not established that this happens, of course, but I suspect that may be one of the early systematic studies we'll do later this year.
4. As for emailing data, that's impossible: the files are gone. As I said, those were exploratory runs looking at sampling frequencies, and once I settled the sampling frequency question to my satisfaction the data files were over-written in subsequent exploratory runs. If anything interesting appears in the current runs I'm doing to look at other configuration issues I'll save the data. Meanwhile, I suggest that Nelson download the program and try it out himself.
RBH
[ 27. July 2003, 03:48: Message edited by: RBH ]
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 27. July 2003 05:00
RBH writes,
quote:
3. There seems to be an impression on Nelson's part that more complex features or processes are necessarily evolved by coopting simpler functions whole. Hence his comment that he finds it "completely outlandish that NOT and OR_N and AND_N is sufficient for EQU, ...". But evolution doesn't have to use a whole simpler function in generating a more complex function. The simpler functions provide organized raw materials for more complex functions. In evolving the simpler functions, chunks of code are organized under selective pressure. The input-output mapping corresponding to NOT requires (at least) 6 instructions to perform. The AND_N mapping requires at least 10. OR_N's mapping requires 7 instructions. Substrings within those longer programs are available for cooption; evolution doesn't have to use the whole simpler function. So the question for Nelson's incredulity is whether the evolved programs that perform the NOT, OR_N, and AND_N mappings contain substrings that are cooptable to perform the EQU mapping.
I know that flagellum-AVIDA comparisons are problematic on many accounts, but you've just pointed out another similarity. When discussing the flagellum, Nelson assumes that "functional subsystems" have to be coopted wholesale (the entire subsystem, with all of its parts, is incorporated) into the flagellum -- as might have occurred with a Type III secretion system, the only subsystem Nelson accepts.
But in fact, once a subsystem exists performing some nonflagellar function, it's perfectly possible for just a part of it (especially if there is duplicated portion of the chromosome containing and extra copy of those genes) to be coopted into the flagellar system; this may be why the flagellum contains homologues to a part of the F1F0-ATPase (FliI) and a part of other secretion systems (lipoprotein L-ring) and a part of ion-powered transport systems (MotAB homologues). Expecting cooption to mean that the flagellum breaks down completely into independent modules [1], where each module is fully functional all by itself, is a highly oversimplified picture.
yersinia
[1] Modular, like Constructicons, for complete lack of a better analogy.
[ 27. July 2003, 05:02: Message edited by: yersinia ]
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 27. July 2003 14:31
yersinia wrote quote:
But in fact, once a subsystem exists performing some nonflagellar function, it's perfectly possible for just a part of it (especially if there is duplicated portion of the chromosome containing and extra copy of those genes) to be coopted into the flagellar system; this may be why the flagellum contains homologues to a part of the F1F0-ATPase (FliI) and a part of other secretion systems (lipoprotein L-ring) and a part of ion-powered transport systems (MotAB homologues). Expecting cooption to mean that the flagellum breaks down completely into independent modules [1], where each module is fully functional all by itself, is a highly oversimplified picture.
That is the level of analysis at which comparisons between biology and simulation are appropriate and informative. The implication on Nelson's part (that he also apparently attributes to Dembski) that one must somehow establish one-to-one mappings of specific structures between the simulation and biological phenomena is a rhetorical device that serves only to misdirect attention to a red herring.
The Avida simulations (and other ALife simulations) can provide analogues of biological processes to inform hypotheses that can be tested in biology. As you say, expecting a complex system to decompose directly into subsystems, each of which is in its current form a fully functional system, is an invalid expectation and serves only to obfuscate what actually happens in both biology and in the Avida simulation. Neither Dembski nor Nelson Alonso seem to understand what cooption means nor what kinds of things can be coopted by evolution. The kind of brick wall construction that they apparently envision for complex structures, where each brick is a discrete inviolable entity, necessarily characterizes neither Avida critters nor biological structures.
RBH
IP: Logged
|
|
|
Topic Closed
Printer-friendly view of this topic
