Why I remain a Darwin Skeptic

Faith & Science Conversation
645

The point is perfectly nested clades are not evidence of evolution if that is all evolutionists look for.

I am interested in evidence for evolution. Perfectly nested clades is held up as the pinnacle of evidence. Yet, as I explained, it cannot be evidence for evolution. If it is the pinnacle of evidence for evolution, then evolution does not have very good evidence. This seems like a problem to me, but perhaps not to professional evolutionary biologists.

FWIW, I’ve also been looking at the data, a collection of 34 mammal mtDNA I have, and off the bat a tree isn’t a very good fit. A created kinds model makes more sense of the data. The reason is the molecular clock implies distance comparison between any three mammals should always form an ultrametric. This requirement is broken for 45% out of all the possible triples, which seems exceedingly high if we have a tree ancestry and a molecular clock.

The reason why the tree necessitates an ultrametric is because all the ancestors are lost to time. We only have descendants in our database. Consequently, the violation of the ultrametric property entails a large number of the ancestors are with us in present time. This doesn’t make sense under the evolution model, since we don’t have any million year old animals living with us today. However, it makes a lot of sense in the created kinds model, where some animals are the archetypes and other animals are derivatives. It also makes sense within Ewert’s dependency graph model. And finally, it is also consistent with Ewert’s dependency graph model if we have triples that do fit the ultrametric property, as can be easily be shown with some set theory.

Thus, the data I have is at best 55% consistent with Darwinian evolution, and 100% consistent with Ewert’s dependency graph. This seems like a win for Ewert’s dependency graph.

646

I look forward to reading your paper in a peer-reviewed biology journal.

3 Likes
647

Even better, I’ll send you a link to the github repo, so you can check the evidence for yourself.

Now, another thing I’ve looked into is full genome clustering, with about 50+ organisms ranging across plants, insects, birds, fish and mammals. Not something that is currently done, AFAIK. The really fascinating thing is that the clustering of the DNA sequences actually matches my intuitive classification based on phenotype. Not a super rigorous result at this point, b/c it’s my subjective taxonomy, but definitely not what I’d expect if Darwinian evolution were true. E.g. I learned in my genetic algorithms classes that there is almost no correlation between genotype and phenotype for complex problems.

648

I do not have the expertise and domain knowledge that comes with a Ph.D. in biology, so; I wouldn’t know how to give it the evaluation it deserves.

1 Like
649

Good point, sorry to push. Thanks for the back and forth, it has been helpful to me, so don’t feel our discussion was in vain. I appreciate the time you put into this site.

650

I think this is where the idea of “perfectly nested clades” entered the thread, and from my limited understanding of biology, that’s not an expectation of evolution.

One reason why is incomplete lineage sorting, as discussed here:

https://biologos.org/articles/series/evolution-basics/incomplete-lineage-sorting

And another way of describing why clades won’t nest perfectly – why branches won’t remain perfectly distinct – comes from @Mervin_Bitikofer in an old thread:

So it seems quite common knowledge that biologists are not expecting a perfect tree. (This was also mentioned repeatedly in the thread @Chris_Falter linked to.) Yes, the general structure is like a tree, like nested clades, but because it is a population-level process, some variation from that is expected, even apart from horizontal gene transfer. That’s why statistics are so integral to biology.

Showing there is not a perfect tree is not really saying anything the biologists haven’t already been saying.

3 Likes
651

Sure, no disagreement there, and that is also not my argument.

But due to all the non responses I’ve received, my best guess is nested clades is not very great evidence for evolution. So, I’ve satisfied my own curiosity on the matter.

652

I don’t pretend to even be able to delve into all your argument to the same depth of expertise (and spent time) that you’ve obviously pursued it, so I cannot give you your desired response, but since I’ve seen others here who do have that expertise responding and I can at least see the overall nature of all your exchanges, I do gain an overall sense of something:

It does seem your thesis, such as it is, finds its only home in a negative or a denial. Something (at least in your estimate) from the positively advanced thesis (evolutionary change through mutation) has not been answered - at least not to your satisfaction. So therefore you are satisfied that your essentially negative thesis is must be the right one, even though you cannot apparently spell out what that thesis even is. Meanwhile those advancing the positive thesis have, not just a pinnacle of evidence (as impressive as that pinnacle is for those of us willing to look at how well it does explain so much!), but an entire wide plateau of stuff from many different fields that all come together and are explained by that positive thesis. So the fact that there certainly remains much to be explained and discovered there (which may or may not even include alleged ‘holes’ you see in it) does not bother the long-time proponents of that thesis so much as what a rival advanced theory that explained things even better would do.

…is a good summary of the situation.

Your curiosity may have been over-easily and prematurely satisfied. Others whose curiosities were more persistent than yours do press on, and - not being satisfied with statements of mere incredulity - they often go on to answer the very problems that were once enshrined by the skeptics as intractable problems.

Hopefully you can see why confidence tends to erode away from the parties whose curiosities are easily satisfied by any apparent silences they may encounter here or there.

What I’m also impressed by, though, is the amount of time and effort (quite the opposite of silence!) that more than one expert here has spent delving into your questions with you while others more driven by their ideological opposition to anything that has a whiff of ID about it would have long ago left you alone with great expanses of silence for you to enjoy as a sign toward whatever conclusions you wished to promote.

But the fact remains, you have pursued questions farther than most of the rest of us here pursue them, and I respect (and look up to you) for that. Your challenges, and the replies they have attracted are very educational to people like me. Thank you for that - and this is not meant to be some conclusive dismissal or ending of it all - quite the contrary. When you feel your curiosity gathering more steam again, it seems that yet more productive engagement could easily follow.

5 Likes
653

In this particular case the question is just a yes or no question: is nested clades evidence of evolution?

If scientists were trying alternative besides just the null hypothesis, then the answer might be yes. But, as best as I can tell from all my readings and exchanges on this site, scientists are only comparing tree vs random.

Now, any dataset that has any sort of discernable structure is almost assuredly going to fit a tree better than randomness. Consequently, the fact a tree fits the data better than randomness tells us nothing interesting, because if any alternate hypothesis were in fact true instead of evolution, we would see the same result.
Namely, that a tree fits the data better than randomness.

My gracious dialogue partners have come up with other novel interpretations of what I am saying, which they’ve aptly debunked. But my actual argumemt remains untouched. Most likely due to poor communication on my part, but I have tried my best to put it clearly in multiple different formats, including a short dialogue, so I’ve probably exhausted my communication abilities in this regard.

I’ve found another thread on this site where a much more knowledgeable participant made the same argument, and encountered the same communication difficulty. Communication is hard even when we can see each other’s body language, so doubly so when all we have is text! The difficulty is very understandable.

Now, what I have noticed in other disciplines is when an argument has been considered before, this sort of miscommunication doesn’t happen. People immediately recognize the argument for what it is, and provide the correct response.

So, what all this miscommunication tells me is that the particular argumemt I am making about the inadequacy of tree vs null testing to tell us anything interesting about the veracity of evolution is either a novel argument never before considered, and thus people have trouble recognizing it, or the argument does not have a good response.

I am doubtful that I’ve thought of anything new that has never occurred to all the professional evolutionary biologists throughout the last century and a half, so I am leaning towards the second option.

654

There is a third option:

We observe today the following stochastic phenomena:

  • inheritance
  • genetic drift
  • natural selection
  • horizontal gene transfer
  • incomplete lineage sorting

In the presence of these phenomena, common ancestry would be expected to produce slightly imperfectly nested clades. Such clades have been observed both narrowly within families and broadly within kingdoms, so those observations are interpreted as a signal of common ancestry.

Common ancestry also provides an excellent fit for many other classes of data (phenotype distribution, fossils, etc.), so common ancestry by evolution has achieved the status of theory.

Think of the explore vs. exploit trade-off (aka one-armed bandit problem) you have studied in CS, Eric. You express frustration that biologists have not abandoned exploitation for the sake of full-bore exploration of new possibilities. But biologists have found a slot machine, evolution, that is yielding new riches every time they pull the arm. And whenever they have experimented with non-evolutionary slot machines (such as Ewert’s model), they’ve come up empty-handed and lost their coins.

So of course they are going to keep exploiting the productive model they have.

Do you desire that they abandon their productive use of evolution? You will have to show them that there are more riches elsewhere. Asking them to abandon the one and only one machine that has been making them fabulously wealthy, and to just have faith that they will find an El Dorado model that will make them even wealthier in knowledge, is asking too much. Those who have chased after an El Dorado on the horizon have discovered mirages. And they do not want to keep chasing mirages.

Show them that an El Dorado is real and robust; then they will come.

Chris

1 Like
655

Can you point me to these alternate models that have been tested on the genetic data?

656

Off the top of my head, here are biologists who have advanced alternatives that have been found wanting:

Ewert
Remine
Sanford
Jeanson

Peace,
Chris

1 Like
657

Hmm, your list confirms what I suspected. Evolutionary biologist find trees because that is all they look for.

658

You have the power to present a better model, or at least to try.

You are still stuck on criticizing biologists for exploiting the one-armed bandit that reliably yields riches. If you want them to explore elsewhere, you’re going to have to show them a more profitable machine.

Criticizing them for exploiting a profitable machine is going to go nowhere fast. Why waste your breath?

Chris

2 Likes
659

I did a couple comments ago, two of them.

But that is tangential to the argument I am making that tree vs null testing tells us nothing about the actual structure of the data, so cannot count as evidence for evolution. Even the usefulness of a tree tells us nothing, because, again a tree will be useful on any structured dataset. The only thing a useful tree says is the dataset is structured, which is consistent with all theories, so distinguishes none of them.

For whatever reason, you prefer to respond to an argument I am not making. So, we probably won’t be making any head way as long as that’s your preference.

Anyways, what I am suggesting scientists must do to provide good evidence for the tree model, I can do myself pretty easily. Just test all the possible graphs for a set of nodes and see what graphs are the best fit, which of those are trees, and which trees fit the evolutionary trees. So I’ll stop wasting everyone’s time as see for myself what the data says.

660

Also, since the nested clade is an invalid argument until I complete my graph analysis, I am afraid you all are not logically allowed to trot it out in support of evolution I’ll have to report any offenders to the logic police >:)

661

How will you do that Eric? There are at least two major problems:

  1. number of possible graphs grows super exponentially with number of nodes
  2. “best fit” is ill defined
662

You dropped a hint a couple comments ago. A real presentation is going to require a whole lot of feedback from biologists, a ton of writing, peer review, the whole 9 yards. Moreover, if you want it to be taken seriously, you are going to need to address the outstanding questions that Ewert has not been able to answer yet. And you are going to have to show how your model is truly parsimonious.

I do not understand how you think you can show some code on Github and then say you have presented a model to the biology community.

But I’m not a biologist. You should be talking to real biologists and asking what they want to see in order for a model to be considered worth looking at.

Peace,
Chris

3 Likes
663

The first problem is particularly vexing. It is at least NP-hard.

Chris

2 Likes
664

Again and again, in fact every time, thanks to the internet facilitating the ultimate marketplace of ideas, I have found every reactionary vestige of unaware un-rational thinking, every unrealised prejudice and blatant ones - sexism, homophobia, race, class all inculcated by so called Christian religion - every Biblicist, textist, undeconstructed, ignorant, magical belief deconstructed. Every time I’ve realised “Is that all you’ve got?”.

So thank you BioLogos. Thank you Chris and T in particular, Steve and Stephen and Mervin. And of course Eric as Chris said, without whose yeah buts this thread would not have let them shine.

1 Like