"I'm treating the mutation rate as a substitution rate" - Dr. Nathaniel Jeanson

Hi, Bill, thanks for the reply. I doubt that’s actually true, but in any case, I was not as specific or clear in my question as I intended, so let me narrow it down.

What responses have you received on this forum about the following claim:

Thanks in advance!

1 Like

You unfortunately have no comprehension of what Joe said. All that’s necessary for all of evolution after the origin of life is changes to existing populations. Nobody here has any clue why you think that’s a problem. Changes to existing populations do include both gene gain and loss, if that’s your objection.

Yes, as the mutations that result in new genes are known, and they’re the sort of thing that happens all the time. You have shown zero understanding of “the Howe diagram”, and you have never even attempted to explain how it supports Behe’s design argument. Also, you should notice that Behe’s ideas are not your ideas of separate creation. You can’t use him to support your basic claims.


Stop invoking the names of people who don’t agree with you in support of your lies. Joe has not acknowledged any agreement anywhere, with any assertion you’ve made about population genetics in this thread.

Stop lying.

And it does.

I agree, which is why the entire community of biology rejects the Intelligent Design just-so story and instead considers the testable theories of common descent and molecular evolution as the only game in town.

Yes. The evidence that shows that novel genes can evolve though mechanisms ranging from duplication and divergence, through recombination, insertion, deletion, and fusion, to de novo by expression of a novel ORF from previously non-coding DNA.

No it actually supports evolution, and common descent as you’ve had explained to you in great detail many times before.


Evolutionists also agree that variation is change within existing populations. That is why there is a nested hierarchy.

Except we are observing more variation then you would expect from reproduction alone. You need a model that explains this variation to challenge Behe’s method.

The models agree on the population size during that time period. Obviously the “when” is exactly what is in dispute.

If universal common ancestry is true, morphology is unreliable. If structural appearance is unreliable, why would we think that molecular appearance is reliable without direct observation of genealogical relationships? Nested hierarchies can obviously be deceiving. I’ve just always thought we should find consistencies in morphology and molecular trees for this to be a compelling argument. If convergent evolution can be appealed to whenever needed, then nested hierarchy has no power as a proof. I don’t know…just thought there must be something more to nested hierarchy for it to be such solid proof of universal common ancestry.

Funny I came across this on another was reading through the other thread - this seems to be a “proof” for common ancestry that really isn’t a good one.

I haven’t thought about it that much. I suppose I posted because I was in the mood for a challenge to think about. :slightly_smiling_face: Can you give me an example of what you’re talking about?

Which theory predicted this? How does a nested hierarchy explain it? Octopus brain and human brain share the same 'jumping genes'

You are only looking at the noise while ignoring the signal. Let’s use this graph as an example:


I would conclude that there is a very strong and statistically significant positive correlation in the data. You would argue that this data is completely unreliable because none of the data points lands on the correlation line.

When we are talking about a statistically significant signal in the data we are referring to the case in the graph above. We don’t expect an absolutely perfect match 100% of the time because there is noise in biological data, just as there is in almost every single scientific data set. What you are ignoring is the signal.

It is a good one, for the reasons given above.

The theory of evolution predicts a nested hierarchy and explains why we observe one. Separate creation does not explain it.


Are you ignoring mutation . . . again?

1 Like

I have? Which limits, relevant here how?

1 Like

It always helps to start with what we agree on. I had to get your final sentence to find it:

I was in essence saying the same thing: statistical significance
Your original question was whether I agreed with this statement:

In that question you did not include a reference to statistical significance, and if you’ll look at my reply, I was pointing that out:

Yes, we agree (at least on this) THAT IS what is important.

And just to re-iterate, when I first explored this topic last year, I was not as familiar with the statistical significance of the NH data. So when I said the following (which is now the third time I’ve said this):

I was referring to the fact that I learned more about what evolutionists seem to “see” in this statistical significance.


HAVING SAID ALL THAT…As I mentioned, I didn’t stop there. I started researching challenges to NH and the ToL. I didn’t go into detail previously, but want to bring this up now: a discussion on what you refer to as “noise”. You had also asked if I agree with this:

That question also needs to be unpacked. It opens up this very critical question: Not just whether it it’s there, but how significant is this “noise”? Within the last few days I actually went back to my notes, and was reminded of the number of challenges there are to NH and the ToL. All which I find others passing off as “acceptable noise”.

So let me ask you this: Can this “noise” also be measured statistically? And if so, what’s the threshold for too much noise? And to get a little more philosophical: Is your noise-tolerance-threshold different than mine?

1 Like

In scientific shorthand, statistical significance is implied when someone says a hypothesis has been supported by data. For example, you can find the line of best fit for any set of data, like these:



However, only one of those data sets will have a statistically significant correlation.

Note: the link for those graphs defines statistical significance at an R-squared of 0.80.

Using the graphs above, noise is the amount of departure from the best fit line.

There are different statistical tests for nested hierarchies. One of those tests is the consistency index, or CI.

And here is figure 1.2.1:

Figure 1.2.1. A plot of the CI values of cladograms versus the number of taxa in the cladograms. CI values are on the y-axis; taxa number are on the x-axis. The 95% confidence limits are shown in light turquoise. All points above and to the right of the turquoise region are statistically significant high CI values. Similarly, all points below and to the left of the turquoise region are statistically significant low values of CI. (reproduced from Klassen et al. 1991, Figure 6).


Consilience of independent phylogenies means you take different but similar parts (loci, positions) of the genome from a collection of organisms, say primates, and then you infer a phylogenetic tree for each of those loci. And then you see that the tree you get for each locus is remarkably similar.

Suppose you pick a particular gene shared between all the species you picked, and infer a phylogeny on the basis of the sequence of this gene found in all these species. You get a particular tree for that gene. Then you pick some other locus in the genome, it can also be a gene, it can be an intron, it can be wherever you want. A gene with a wildly different function, or maybe even a pseudogene that is nonfunctional. You again infer a phylogeny on the basis of this new gene and you get a new tree from this.

Now you compare the first tree to the second tree. They’re virtually identical.

So now you ask yourself: Why do you get the same tree from these different loci? The gene tree you get from one locus, which has a particular function unrelated to the function of the other locus, is still extremely similar. Why? Why are they extremely similar?

Predicted what? You’ve got the logic inverted here. A nested hierarchy is what we are demanding an explanation for (and we are explaining it’s existence with the theory of common descent). The nested hierarchy is not itself proposed to explain something else.

The nested hierarchy is an extremely strong pattern in the data we are noticing, and we are demanding an alternative, testable, and superior explanation for it from people who propose to reject the theory of common descent.



I tried to find the discussion but maybe it is too old. The discussion was with if the current state of population genetics mathematics assumed the existence of a population. I believe your answer was that this is true today.

The relevance is if current population genetics mathematics could model the origin of the gene patterns in the Howe diagram.

Ehh, no. It is less reliable than DNA is, but that doesn’t make it unreliable. That’s just such a prototypical example of dichotomous thinking.

Scientists can test phylogenetic methods on known phylogenies (such as the known history of viruses evolving here and now, or our own family histories, and so on). If the methods they use to construct phylogenies from DNA sequences very accurately reproduces the known genealogical histories (and they do), that gives us reason to have confidence in their reliability when we don’t know the real genealogical history.

And scientists can do computer simulations of evolution using many different parameters and assumptions to determine how, when, where, and to what degree these factors result in erroneous phylogenetic trees, and in turn use these to understand phylogenies derived from real biological data to see if these show evidence of having undergone events or mechanisms that produce these errors.

And they can do various statistical tests on trees to see how consistent they are (such as the bootstrap method).

Can be =/= always are.
Can be =/= are to an absolute degree.

Again with dichotomous thinking. Theobald wrote about this all the way back in 1999:

In science, independent measurements of theoretical values are never exact. When inferring any value (such as a physical constant like the charge of the electron, the mass of the proton, or the speed of light) some error always exists in the measurement, and all independent measurements are incongruent to some extent. Of course, the true value of something is never known for certain in science—all we have are measurements that we hope approximate the true value. Scientifically, then, the important relevant questions are “When comparing two measurements, how much of a discrepancy does it take to be a problem?” and “How close must the measurements be in order to give a strong confirmation?” Scientists answer these questions quantitatively with probability and statistics (Box 1978; Fisher 1990; Wadsworth 1997). To be scientifically rigorous we require statistical significance. Some measurements of a given value match with statistical significance (good), and some do not (bad), even though no measurements match exactly (reality).

You can in principle have twenty ultra sensitive thermometers that can measure temperature to one part in a billion, all disagree on the exact value for the temperature in the same room, yet all of them can be within a range of one part in a thousandth of a degree. That would be a remarkable degree of corroboration even if the temperature output by each thermometer is different from every other.

We do find consistency, but you have to understand that it comes in degrees, and in part because anatomical convergence can happen due to both natural selection and certain forms of developmental constraint (which can therefore make trees derived from morphology alone misleading) have biologists largely moved on to molecular phylogenies instead and are testing tree consistency by comparing different molecular trees to each other, rather than by comparing morphology with molecular trees.

In other words the problem of convergence at the level of gross anatomy as a confounding factor in phylogenetics is recognized, not ignored, and it is addressed by other methods of measuring tree consistency using different molecular loci and traits where convergence becomes exceptionally statistically unlikely.

Well it can’t, of course, because in principle at a certain ubiquitous level of convergence the tree disappears entirely leaving you without nesting sets at all. And then there’d be no nested hierarchy, and then it wouldn’t make sense to invoke convergence. You only invoke convergence when something deviates from the main nesting pattern. If convergence was ubiquitous, there would be no main nesting pattern.

But there is still a nested hierarchy, even if convergence does happen in some clades. Despite living in the ocean and sharing a certain degree of gross overall body shape, no phylogenetic method confuses whales for sharks. Or mammals for birds. Or plants for bacteria. Just to pick some more obvious examples.
Phylogenies derived on the basis of morphology can be difficult within certain clades because the number of characters from which to derive the phylogeny(usually a few hundred at best) is relatively small compared to molecular data, that can use many tens of thousands.

You have to understand that consistency of the nested hierarchy goes far beyond a comparison of morphology to DNA too. You can compare different parts of the genome as explained in my previous post. You can compare mitochondrial to nuclear gene trees. Trees from genes on different chromosomes. Tree from coding to non-coding DNA trees. Enzymes to developmental genes. Introns to retrovirus-derived retrotransposons. Pseudogenes to centromeres. And so on ad infinitum.

Have you considered that you actually have no idea of the degree of consistency you get?


So they don’t agree at all. That there may be two quite different time periods for which both models predict the same population is not agreement at all. And let us point out that Jeanson’s time scale doesn’t match the well calibrated and supported time scale of human history. Any theory that is forced to posit a young earth is dead from the start.

Just not the case. There are problems with morphological character analysis that make it less accurate than molecular analyses, notably the great imbalance in the quantity of data available. But given that, the match is quite good, and discrepancies are rarer than you imagine, especially when there’s enough data available for both data types. Another problem is that scoring morphological characters is inherently more subjective than scoring sequence characters, for which there are only four discrete possible states.

We usually do. But as others have mentioned, consistency in sequence data from different parts of the same genome is an equally valid test.

There is much more than you know. And it’s not just for universal common ancestry. It also works within any subsets of life that you care to name, and creationists need to explain all of it but can explain none of it.

That’s not quite a sentence. You need to rewrite to make it comprehensible. But it’s better evidence (“proof” isn’t a word we use here) than you know.

Literally hundreds. Choose any gene you like and compare it across ape genomes, and you will find the same tree is by far the most frequent. (Some genes will differ because of incomplete lineage sorting, a well understood phenomenon, but even the fraction of different trees is predictable.) “Ape”, of course, includes humans.

There’s nothing in that piece that’s relevant to a nested hierarchy, just the presence of some LINEs. Most species have LINEs.


Yes, it can. You’ve been told how dozens of times. Why do you insist on pretending you haven’t been given an explanation?


Huh? Maybe you misunderstand…or I don’t get your point.

We have independent historical population size data. Jeanson says his model matches the population growth data from 1000 BC to the present. That’s in spite of the fact that many of the migrations the evolutionary model says are prehistory actually happen sometime in the last 3000 years in his model.

Sure, I’ll consider it. I’d like to read a paper or two on what you’re describing. Share your favorites…


A comprehensive molecular phylogeny based on 34,927 bp (after correction for ambiguous sites from the original dataset of 43,493 bp per operational taxonomic unit, OTU) amplified from 54 nuclear genes in 191 taxa including 186 primates representing 61 genera is presented (Figure 1, Figure 2, Figure S1, Table S1, and Table S2). The phylogeny is highly resolved, with bootstrap values of 90–100% and Bayesian posterior probabilities of 0.9–1.0 at 166 of the 189 nodes (88%)(Table 1, Table 2, Table 3). Further, only 3 of 189 nodes (nodes 28, 38, 158) are polytomies in the bootstrap analyses (Table 1 and Table 3; Figure 2, Figure S1). (Note: nodes listed hereafter refer to Figure 2, Figure S1, Table 1, Table 2, Table 3). Roughly equal amounts of coding (14742 bp) and non-coding (17185 bp) genomic regions were sampled from X chromosome (4870 bp), Y chromosome (2630 bp) and autosomes (27427 bp) (Table 4) using newly developed PCR primers derived from a bioinformatics approach specific to primates in addition to primers from previous large-scale phylogenetic analyses (Materials and Methods, Tables S2, S3, S4).

Please understand that for 191 species approximately 6.21 x 10407 different rooted phylogenetic trees are possible. That means there are an incredible number of ways to make trees, so it is unfathomably unlikely you should get extremely similar ones over and over again when the sequences from different genetic loci obviously don’t constrain each other to produce similar trees(why would they?).
And yet the trees come out virtually identical every time when resampled across 35 thousand basepairs of DNA coming from 54 different genomic loci with wildly different functions. So why do we get this degree of consistency among the trees?

It strains credulity to say it’s just a happy accident, but common descent predicts it (and therefore explains it), so what is the creationist explanation and does it actually make sense?


Shouldn’t that be “in spite of the fact that many of the migrations the archaeological evidence says are prehistoric happen sometime in the last 3000 years in Jeanson’s model”. That certainly seems to be the case with both haplogroups R1b in Europe and E1b1 in the Middle East.


If you’re using the first person there, I expect you to describe the data that you have analyzed.

That’s just hearsay from a nonexpert in the field. What does the evidence tell you?

1 Like