Uncommon Descent Serving The Intelligent Design Community

The Elephant in the Room

Share
Facebook
Twitter
LinkedIn
Flipboard
Print
Email

We are regularly told by proponents of evolutionary theory, from Darwin right up to the present day, that purely natural processes, such as random mutations and natural selection, have the ability to build, construct, fashion, purpose and create remarkable machines. Machines that rival, and in many cases surpass, our most advanced technologies.

We are assured in no uncertain terms that such natural processes have this great creative power. Yet when examples are sought, we are invariably given examples that either did not come about through purely natural processes (see Berra’s Blunder), or examples that are trivial in scope. But nothing that even comes close to verifying the grand claims of the evolutionary creation story.

There is a huge elephant in the room.

Why, if evolutionary processes are so incredibly adept at producing remarkable technologies that surpass our capabilities, do we not see such evolutionary processes being put to good use on a regular basis?

All around the world, every day, millions upon millions of new inventions, designs, projects, programs, and other creations are being pursued. Yet the most awesome creative force of all, so we are assured, is for some reason notably absent. Occasionally someone will claim that evolutionary processes were responsible for creating this or that product (the NASA antenna being the example most often trotted out, even though it is not a proper example of purely natural evolutionary processes). Sometimes someone will assert that an “evolutionary algorithm” has produced something mildly interesting (like the questionable and potentially flawed Avida results touted several years ago in Nature). But by and large, this alleged remarkable creative force is absent, irrelevant, a “no show,” when it comes to actually creating things in the real world.

Now the evolutionary proponent will no doubt argue that the reason is simple: not enough time. Easily impressed with all the zeroes in a number like the billions of years of Earth’s history, the evolutionist reposes faith in the power of deep time to take what is clearly an impotent process in the short term and turn it into the most potent creative force in the long term. But when the actual numbers are reviewed and the actual requirements for construction of functional creations assessed, it becomes clear that those zeroes in the age of the Earth or even the age of the universe are but a rounding error and are unhelpful in addressing the larger issue.

To be sure, a trial-and-error process like random mutations and natural selection can occasionally do something interesting – if there is a large enough population and a strong enough selective pressure. Behe has spent time searching for this “edge of evolution,” while in stark contrast most evolutionists never even bother thinking about what evolutionary processes can actually accomplish in the real world, simply taking it as an article of faith that “with evolution nothing is impossible.”

More to the point, such minor changes even when they do show up do not constitute evidence for the larger evolutionary claims. Particularly when many of the alleged examples of evolution’s power turn out to be, on closer examination, examples of breaking a machine, rather than building it.

So the elephant in the room remains. Design is a critical aspect of our modern lives. Design occurs across the spectrum of disciplines and across the globe on a near constant basis. Yet the most potent creative force that allegedly ever existed, that of evolutionary mechanisms, is noticeable in its near complete absence – dabbling at the fringes, only occasionally participating, rarely influencing, never doing much of any real consequence.

We might be forgiven for wondering if perhaps this is all the evolutionary mechanisms have to contribute.

Or all that they ever did.

Comments
Zachriel @357:
Continuity. Even with simple mutation, a non-reproducing entity will almost inevitably go extinct before exploring much of the search space.
Agreed. So reproduction increases the amount of the search space that can be explored before extinction. Which means we are increasing our probabilistic resources brought to the table.
Recombination within a population allows exploration of areas of the search space not available to mutation alone. We’ve pointed this out before, but your latest comment indicates you have yet to digest it. A simple example was provided above.
Nope. You haven't explained any such thing. You've pointed to recombination as though it does something intrinsically different. Yet all recombination does is quickly bring together two areas of the search space that were explored by two different organisms/lineages. You have not provided a single reason to think that such search space would be "unavailable" to simple mutation. It is just a question of time and resources. So, again, we are back to the basic point that recombination arguably provides a chance to explore a larger portion of the search space within the time allotted. I don't know, maybe we are saying the same thing. Are you just saying that recombination allows us to explore part of the search space that would be unavailable to simple mutation in the same amount of time? Or that that part of the search space would be unavailable to simple mutation ever? The first is correct. The second isn't.Eric Anderson
March 12, 2015
March
03
Mar
12
12
2015
09:54 AM
9
09
54
AM
PDT
"Design is a mechanism, too. And we have evidence for intelligent design in biology. OTOH your position has nothing but promissory notes." No, design is the claim. You actually have to have a model to test itCHartsil
March 5, 2015
March
03
Mar
5
05
2015
09:39 AM
9
09
39
AM
PDT
wd400:
You will probably win the lottery several times before a non-reproduction version hits the target whereas it will takes seconds with reproduction.
All the reproduction in the universe will not help unguided evolution reach any biological targets that involve multiple protein interactions.Joe
March 5, 2015
March
03
Mar
5
05
2015
09:23 AM
9
09
23
AM
PDT
Design is a mechanism, too. And we have evidence for intelligent design in biology. OTOH your position has nothing but promissory notes.Joe
March 5, 2015
March
03
Mar
5
05
2015
09:21 AM
9
09
21
AM
PDT
"Design is a mechanism. And ID’s specifics come in the design detection methodology." No, design is the claim. You have to actually evidence itCHartsil
March 5, 2015
March
03
Mar
5
05
2015
09:16 AM
9
09
16
AM
PDT
I am focusing on a very nuanced issue: what does reproduction itself bring to the table. It is a question that it seems most evolutionary proponents never stop to take time to think through carefully, because it is just assumed to be such a fundamental aspect of evolution that no-one would ever even think of asking the question. Nah, it's a question evolutionary biologists learn the answer to very early on. Reproductive excess (that there are more possible organisms than an environment can carry) does increase the "probabilistic resources" available (if that phrase can be construed to mean something). But the more important consquence is that "good alleles can outlive their host. Members of a population today start out with all of the benifiests of many previous rounds of selection. You could try reading The Blindwatchmakerif this is still not clear to you. Or, as I have suggested, make your own "weasel" program. You will probably win the lottery several times before a non-reproduction version hits the target whereas it will takes seconds with reproduction.wd400
March 5, 2015
March
03
Mar
5
05
2015
09:08 AM
9
09
08
AM
PDT
Design is a mechanism. And ID's specifics come in the design detection methodology. OTOH your position doesn't have any specifics, doesn't have a methodology and can't even be tested.Joe
March 5, 2015
March
03
Mar
5
05
2015
09:06 AM
9
09
06
AM
PDT
"Unguided evolution cannot account for biological reproduction. And recombination is a mutation." So exactly which step is the result of exactly what design mechanism? Be specific.CHartsil
March 5, 2015
March
03
Mar
5
05
2015
09:02 AM
9
09
02
AM
PDT
Unguided evolution cannot account for biological reproduction. And recombination is a mutation.Joe
March 5, 2015
March
03
Mar
5
05
2015
08:59 AM
8
08
59
AM
PDT
Eric Anderson: what does reproduction itself bring to the table Continuity. Even with simple mutation, a non-reproducing entity will almost inevitably go extinct before exploring much of the search space. The time to extinction can be calculated based on the landscape. For instance, if 1% of mutations are fatal, then the chance of survival is tiny after a thousand generations, and negligible after several thousand generations. With reasonable fecundity, the line can continue indefinitely. Eric Anderson: All of the answers that have been provided (you, Zachriel, Piotr, Me_Think) — answers which are perfectly valid, by the way, and most of which I agree with — boil down to one thing: additional probabilistic resources. That is incorrect. Recombination within a population allows exploration of areas of the search space not available to mutation alone. We've pointed this out before, but your latest comment indicates you have yet to digest it. A simple example was provided above.Zachriel
March 5, 2015
March
03
Mar
5
05
2015
08:56 AM
8
08
56
AM
PDT
wd400 @354:
I really don’t think it can be explained more simply than it already has been.
We have probably gotten where we need to in the discussion. I am focusing on a very nuanced issue: what does reproduction itself bring to the table. It is a question that it seems most evolutionary proponents never stop to take time to think through carefully, because it is just assumed to be such a fundamental aspect of evolution that no-one would ever even think of asking the question. All of the answers that have been provided (you, Zachriel, Piotr, Me_Think) -- answers which are perfectly valid, by the way, and most of which I agree with -- boil down to one thing: additional probabilistic resources. So far without exception they have been subsets of that issue. It is important to recognize that this is the case. Again, let me be clear that I completely agree that a (substantial) population will have a much greater likelihood of stumbling across a favorable mutation (or other advantageous evolutionary change) than will a single lineage. In most cases by several orders of magnitude. Whether a population in fact stumbling across such a favorable change is now likely or plausible depends on the numbers. Whether that change will persevere and become fixed in the population is yet an additional question. But what reproduction (and a large population) ultimately brings to the table in the blind search for advantageous change is additional resources.Eric Anderson
March 5, 2015
March
03
Mar
5
05
2015
08:46 AM
8
08
46
AM
PDT
Piotr @338:
The problem here is that such a population is inevitably doomed. Sooner or later either a fatal accident will befall its sole representative before it produces its single descendant, or enough deleterious mutations will accumulate to kill it. In a viable population harmful mutations are kept in check by natural selection, but in a one-member population the effect of natural selection is instant extinction.
Agreed. A (somewhat) large population protects the species against the nearly-inevitable destruction by mutation or the hazards of nature that would befall a single lineage. Thus, a population makes the likelihood of surviving, and therefore potentially finding, a new functional evolutionary change greater.Eric Anderson
March 5, 2015
March
03
Mar
5
05
2015
08:26 AM
8
08
26
AM
PDT
EA @342, I really don't think it can be explained more simply than it already has been.wd400
March 5, 2015
March
03
Mar
5
05
2015
08:22 AM
8
08
22
AM
PDT
Unguided evolution cannot account for recombination.Joe
March 5, 2015
March
03
Mar
5
05
2015
08:21 AM
8
08
21
AM
PDT
Eric Anderson: One population might have a mutation that helps with environment x, while another might have a mutation that helps with environment y. And because we now have two populations they somehow navigate differently through the landscape to find a mutation that will help with environment z? Yes. Recombination can explores areas of the landscape otherwise unavailable. A simple example above was provided above. gpuccio: Recombination can only work if modular function is present. Without recombination, word evolution is very limited, and gets stuck on local maxima. With recombination, most the dictionary becomes available. gpuccio: So, you can easily show us how to go from SUBJECTIVISM to EXTEMPORIZER through one letter variation and functional intermediates. Twelve-letter words easily evolve if you include a population with recombination.Zachriel
March 5, 2015
March
03
Mar
5
05
2015
05:34 AM
5
05
34
AM
PDT
#350 gpuccio, I don't think their model implements quasi-neutrality (where fitness is allowed to vary minimally within some small threshold). The paper by Kouyos et al. (2012; see above) shows that for a threshold of 10^(-3) a nearly neutral walk in a rugged landscape can reach a distance of about 100 mutations. By the way, there are other reasons why the study is not very realistic in its simplicity. For example, it assumes a fixed, stationary landscape, while real fitness landscapes can and do change, tracing changes in the environment (in this case, the host, Escherichia coli keeps evolving as well), and forcing the phage to re-adapt to any changes that affect its reproductive success. Perhaps more importantly, the authors start with a randomly generated polypeptide. In natural conditions, any starting sequence would in all likelihood already confer a higher-than-random fitness -- or, in the spirit of Wagner, would be fine-tuned for some other functions with a potential for exaptation via minimal nearly neutral modifications. The evolution of new functions does not start in a "fitness vacuum".Piotr
March 5, 2015
March
03
Mar
5
05
2015
04:49 AM
4
04
49
AM
PDT
Piotr: By the way, I don't agree that the paper does not evaluate neutral or quasi neutral walks: "Although each sequence at the foot has the potential for evolution, adaptive walking may cease above a relative fitness of 0.4 due to mutation-selection-drift balance or trapping by local optima." Emphasis added.gpuccio
March 5, 2015
March
03
Mar
5
05
2015
03:49 AM
3
03
49
AM
PDT
Piotr: We are in no hurry at all! I am very busy too. :)gpuccio
March 5, 2015
March
03
Mar
5
05
2015
03:45 AM
3
03
45
AM
PDT
If rivers evolve then so do individuals.Joe
March 5, 2015
March
03
Mar
5
05
2015
03:24 AM
3
03
24
AM
PDT
#354 gpuccio,
Of course, you are free to dream on, even when evidence is against your theory. The paper estimates a starting library of 10^70 sequences to reach the wild type by mutations. That is not realistic, even with “vast populations, millions of generations”.
The estimate is based on a simplified theoretical landscape model which implements ruggedness but doesn't implement (quasi-)neutrality or any type of non-point mutations (to begin with); it should therefore be taken with a rather large grain of salt. Phages are arguably the most numerous "life forms" on earth, easily outnumbering their bacterial hosts by an order of magnitude. A coliphage can repeat its replication cycle many times a day, producing dozens if not hundreds of copies per lytic cycle. Exploiting the brute force of large numbers is something that phages must be very good at.
The point is that after a short time the landscape just prevents the traversing to a very specific, isolated local optimum. IOWs, the wildtype optimum, the very efficient optimum we observe in reality, is a tiny island in the ocean, and cannot be realistically reached.
Rather than that, it's one in a large archipelago of such islands which are isolated from one another if you consider only single point mutations and require every step to be advantageous (ignoring nearly-neutral walks).
The supposed role of recombination is just wishful thinking: they had to say something, of course. It is not ignored “for simplicity”, but because the paper can offer exactly nothing about that hoped future explanations, because they have exactly nothing. Since when is the lack of any evidence for an argument called “simplicity”?
The objective of the study was to see if random point mutations plus selection were enough for the phage to regain a fitness index comparable with that of the wild type. If they don't seem to suffice, some of the simplifying assumptions are probably unrealistic. That's what the authors really conclude. It wasn't their purpose to come up with a more sophisticated model, but such models are being developed: http://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1002551
Again, recombination is just a form of random variation. It tests states, exactly like SNP, deletion, insertion, frameshift mutation, and so on. A single event is a single event, and a single new state is tested, either it differs from the previous of one AA, or of all. The only difference in recombination is that it can retain in some measure existing functional sequences. By the way, it is the simplest mechanism to be tested: homologies are there for that purpose.
If your search space is the DNA sequence space, point mutations differ radically from all types of DNA shuffling, and even from simultaneous double mutations: they can only cover one Hamming distance in each step. It's a very serious limitation in a highly rugged landscape.
That said, I must say again that I appreciate your discourse. You are trying to make reasonable arguments, and even if I don’t agree with them, it is refreshing for me. However, I still can’t see any “connection” ( :) ) between your arguments (which, in essence, are about the topology of functional proteins in the search space, an old an important controversial issue in our discussions here) and Wagner’s ideas. Am I missing something?
Thank you for your kind words. I'm slowly getting to the point, as regards Wagner's ideas. I hope we are not in a hurry (I'm having a hectic time at work).Piotr
March 5, 2015
March
03
Mar
5
05
2015
03:05 AM
3
03
05
AM
PDT
gpuccio @ 344,
So, you can easily show us how to go from SUBJECTIVISM to EXTEMPORIZER through one letter variation and functional intermediates. Piece of cake for you.
Don't you think I tried ? :-) There is no pathway (Although if you calculate the Levenshtein distance or Edit distance between your 2 strings, the value obtained is 11)Me_Think
March 5, 2015
March
03
Mar
5
05
2015
12:53 AM
12
12
53
AM
PDT
Piotr:
The wild type is the product of vast populations, millions of generations, and a greater variety of mutation types (ignored in the study for simplicity, as the authors say). It can certainly achieve much more fine-tuning than in-vitro experiments. Relative fitness lower than that of the wild type does not mean “no infectivity”; it means lower infectivity. The mutants would be outcompeted in the wild, which is hardly surprising. The descendants of the successful survivors of innumerable rounds of selection must be tough.
Of course, you are free to dream on, even when evidence is against your theory. The paper estimates a starting library of 10^70 sequences to reach the wild type by mutations. That is not realistic, even with "vast populations, millions of generations". The point is that after a short time the landscape just prevents the traversing to a very specific, isolated local optimum. IOWs, the wildtype optimum, the very efficient optimum we observe in reality, is a tiny island in the ocean, and cannot be realistically reached. The supposed role of recombination is just wishful thinking: they had to say something, of course. It is not ignored "for simplicity", but because the paper can offer exactly nothing about that hoped future explanations, because they have exactly nothing. Since when is the lack of any evidence for an argument called "simplicity"? Again, recombination is just a form of random variation. It tests states, exactly like SNP, deletion, insertion, frameshift mutation, and so on. A single event is a single event, and a single new state is tested, either it differs from the previous of one AA, or of all. The only difference in recombination is that it can retain in some measure existing functional sequences. By the way, it is the simplest mechanism to be tested: homologies are there for that purpose. That said, I must say again that I appreciate your discourse. You are trying to make reasonable arguments, and even if I don't agree with them, it is refreshing for me. However, I still can't see any "connection" ( :) ) between your arguments (which, in essence, are about the topology of functional proteins in the search space, an old an important controversial issue in our discussions here) and Wagner's ideas. Am I missing something?gpuccio
March 5, 2015
March
03
Mar
5
05
2015
12:27 AM
12
12
27
AM
PDT
Me_Think at #335: So, you can easily show us how to go from SUBJECTIVISM to EXTEMPORIZER through one letter variation and functional intermediates. Piece of cake for you.gpuccio
March 5, 2015
March
03
Mar
5
05
2015
12:13 AM
12
12
13
AM
PDT
Zachriel: "Why? Because they only used point mutation. Recombination can allow the search to move between local maxima." I don't agree. Recombination can only work if modular function is present. And recombination has a huge search space too. And recombination can be detected by homology. IOWs, 3 good motives not to believe in recombination as the explanation for complex functional information in biology. Beware, I am not denying that recombination can have some role. It is, in itself, one of the main modalities of design (modular design, object oriented programming). But, as an attempt at explaining biological information by RV + NS, it is really a fairy tale. "In any case, the experiment shows that a random sequence can have function, and that the degree of function can be substantially increased through natural selection." I agree. That's why I say that the rugged landscape paper is a good paper. It shows many interesting things, and it is a fair attempt (with some limitations) to establish the powers and limits of RV and NS in a context where they are extremely favored.gpuccio
March 5, 2015
March
03
Mar
5
05
2015
12:10 AM
12
12
10
AM
PDT
wd400 @341:
No. It will take many orders of magnitude fewer “opportunities” for the target to be achieved.
What does this mean? Are you saying that the search landscape changes with a larger population?
But more importantly, reproduction allows (future) populations the advantage o individually-improbably “good” mutations. That makes navigation through the fitness landscape a very different prospect.
Can you explain this a bit more? One population might have a mutation that helps with environment x, while another might have a mutation that helps with environment y. And because we now have two populations they somehow navigate differently through the landscape to find a mutation that will help with environment z?Eric Anderson
March 4, 2015
March
03
Mar
4
04
2015
08:32 PM
8
08
32
PM
PDT
Why? Because with reproduction it would have more opportunities to find the target?
No. It will take many orders of magnitude fewer "opportunities" for the target to be achieved. Methink, That the gene pool is larger is only a small part of importance of reproduction. "reproductive excess" does mean more of the fitness landscape is explored in a given generation. But more importantly, reproduction allows (future) populations the advantage o individually-improbably "good" mutations. That makes navigation through the fitness landscape a very different prospect.wd400
March 3, 2015
March
03
Mar
3
03
2015
07:38 PM
7
07
38
PM
PDT
Eric Anderson @ 339 Reproduction increases the available gene pool, which means a bigger genotype network, which in turn means more opportunity to find a new phenotype - the higher the number of vertices, the higher will be the hyperdimensions ( Number of vertices = 2^d, d = dimensions) of the network so it is easier to find a new phenotype.Me_Think
March 3, 2015
March
03
Mar
3
03
2015
07:10 PM
7
07
10
PM
PDT
wd400:
As I said earlier. If you doubt that reproducton/selection are important to evolutionary processes you ought to code up a “weasel” program with and without reproduction and see how long each takes to find the target.
Why? Because with reproduction it would have more opportunities to find the target?Eric Anderson
March 3, 2015
March
03
Mar
3
03
2015
05:08 PM
5
05
08
PM
PDT
#337 One can treat one individual as a degenerate population. The problem here is that such a population is inevitably doomed. Sooner or later either a fatal accident will befall its sole representative before it produces its single descendant, or enough deleterious mutations will accumulate to kill it. In a viable population harmful mutations are kept in check by natural selection, but in a one-member population the effect of natural selection is instant extinction.Piotr
March 3, 2015
March
03
Mar
3
03
2015
11:43 AM
11
11
43
AM
PDT
* For wd400?s benefit: again, talking about a single organism that reproduces asexually once and then dies; no large population ever exists OK, but you were claiming an individual evolves. In this highly contrived case you would have a lineage not an individual. The wider point still seems to be a distraction. Of course mutation+reproduction+selection "only" increases the probability of a given outcome compare to mutation alone. What else could it do? What outcome would be impossible to achieve my mutation alone, or indeed by atoms hitting each other at random? Increasing the probability is the only show in town. EA: I must have missed this earlier in the discussion, but ran across it now. It sounds like we are in agreement that the key thing brought to the table is additional probabilistic resources. As far as evolution is concerned, the fundamental, key aspect of reproduction/recombination/and so on, is helping to increase the probabilities. As I said earlier. If you doubt that reproducton/selection are important to evolutionary processes you ought to code up a "weasel" program with and without reproduction and see how long each takes to find the target.wd400
March 3, 2015
March
03
Mar
3
03
2015
10:00 AM
10
10
00
AM
PDT
1 2 3 13

Leave a Reply