Ashwin on Common Descent

(Ashwin S) #21

I would define it a more measurable way… perhaps in terms of functional capability… I.e human beings are capable of far more things than apes… This gap is smaller between mice and rat (and perhaps any other animal).

I agree provided the rule does not involve abandoning mathematical reasoning at any point without verifiable reasons. We need to be able to verify the correlation has something to do with speciation. I can think of a test that might work… Though I don’t know if anyone has done it.

I understand this… though u will not confine myself to the basic assumption of natural causes… Because I don’t see the assumption as true in all cases (I.e I believe God can intervene miraculously).

(S. Joshua Swamidass) #22

Great, we agree.

That is exactly correct. It might change in the future, but no one has done it yet. Though, I have proposed one model that might work, but we can get to that later.

Fine. Do that, but ONLY if you make that leap outside of science. I’ve already granted that can be legitimate outside of science. The case for common descent is only about what we can say in science in the language of science. Take into account God’s action (outside of science) and all bets are off. Maybe common descent is false, but somehow God created in a way that it really looks like common descent to scientists.

Reasons to Believe is an OEC organization, with which @AJRoberts works, and I am sure she would agree to those rules. @Agauger can comment here too, though I think she would play by those rules too. If you keep the inference to special creation out of science, as believe the two of them are, then you are on solid ground. That is one reason why I am supportive of their efforts.

Okay, so next part of the case next.

(S. Joshua Swamidass) #23
  1. The functional/design difference between humans and chimps is GREATER THAN the functional/design difference between mice and rats.

However, this leads us a surprising fact of nature…

  1. The genetic difference between humans and chimps is LESS THAN the genetic difference between mice and rats.

Notice, this is NOT an argument that reduces to “humans and chimps are sooooo similar!” No, this is a comparative statement, for which I have not even specified how we are measuring genetic differences. Really, any generic genome-wide measure of differences will do. They will all give the same relative answer. This is a robust finding.

Which then brings us to the big question…

  1. Why is it that #1 did not correctly predict #2? Why is it that the large gap in functional/design differences between humans and chimps did not translate into a larger difference in genetic differences (comparative to mice and rats)?

That is the fundamental puzzle of nature. Evolutionary science, making use of common descent, provides a mathematical answer to this puzzle.

  1. It tells us that the formula D=TR should approximately hold for genetic differences. The genetic data does, in fact, follow this approximate pattern. For many types of differences, we can measure D, T, and R independently and it matches this formula.

  2. The formula D=TR does not explain functional/design differences, which do not work by this formula at all, and can change at much different rates.

That is the point of the argument. The genetic features change in a regular predictable way, but the design/functional features do not change in a regular predictable way. That is why there can be such a large discrepancy between the two, as we see in this case.

  1. So this explains why we see humans/chimps so similar genetically, but functionally so different, and mice/rats the opposite. No other known design principle explains this curious feature of the world, and many many others like it.

So that is why we take common descent to be a rock solid finding of genetic science. That is also why the difficulty in mapping phenotype/forum/function to genetics does not undermine, but actually bolsters, the argument. Phenotype/forum/function does not change at a regular rate, but genetics does. That is why we see the discrepancy. That is why it scientifically looks like common descent. Common descent is the plain reading of genomes.

Now, it is legitimate to wonder if God intervened at points in a process of common descent to make special species from existing species. Maybe that is what happened. There is no strong evidence for or against this. Practically speaking, this is no different than inspiring mutations in common descent, or “god guided” evolution as @gbrooks9 likes to call it. You can certainly take that path.

The EvoGrad Blogger
(George) #24

An excellent conclusion!

Smaller changes in genetics can lead to big changes in phenotype or in behavior!


We should reference the classic Mary-Claire King & Alan Jukes paper, “Evolution at two levels in humans and chimpanzees” (Science 11 Apr 1975: Vol. 188, Issue 4184, pp. 107-116
DOI: 10.1126/science.1090005)

Take home message: Regulatory mechanisms seem to be more plastic to change.

Aside: After her Ph.D. work with Wilson, King went on to discover BRCA1, a mutation involved in some forms of hereditary breast cancers.

(S. Joshua Swamidass) #26

That’s right. As I explained…

(Ashwin S) #27

Isn’t this just another way of saying we are measuring the wrong stuff? Ultimately scientists are measuring the difference in nucleotides in the chain… this does not say anything about the functional significance of the change. And that might not depend on the arrangement of nucleotides. We know the three dimensional shape of DNA plays a huge role.
Why doesn’t this fact make scientists question whether they looking at DNA and change it causes in the right way? Wouldn’t that be a valid question?

Isn’t this saying that genetic changes are not correlated to function/design. How does this prove anything other than our ignorance?

You yourself said we don’t know how change in genes relate to change in function. So what exactly does common descent explain? That it’s OK for genes to vary over time?

If we view mutation as an error that occurs during sexual reproduction . There is nothing surprising about there being a correlation in the mutation rate existing within species which have similar reproductive mechanisms. This would be expected.

And if we assume speciation is saltational, the vast majority of the time, the genome is influenced by drift which is influenced by how DNA is handled during sexual reproduction.
My intuition is D=TR is not influenced by common descent. It depends on mode of reproduction… for example, the mutation rates for organisms which reproduce asexually will not be correlated with organisms that reproduce sexually.

The only thing this paradox emphasises to me is that we have no overall idea how DNA leads to form/behaviour.

Disclaimer: Pls understand, I am honestly trying to process your argument. I am not being intentionally dense.

(S. Joshua Swamidass) #28

Read it over again. Your objections are not objections. They are literally part and parcel of the argument.

No. There is a formula that predicts it. That means that it is the right stuff to look at to test the formula. Looking at other stuff is the wrong thing to do.

This is a VERY robust finding that is not tightly dependent on the specific way we look at DNA. Notice, I did not even specific how we are doing here except to say some plausible genome-wide measure of difference. As long as we measure it the same way in both pairs, we will get the saem result.

Look at the logic of the argument. This explains why we can get one answer genetically, and another by function.

You are looking at the wrong data. The formula does not predict that rate of change. It predicts something else. So it is not relevant to what the formula predicts.

What is surprising is that common descent and D=TR predicts the vast majority of the data it is designed to predict.

Saltation is usually about function, not about genetics in this case. Once again, we are making no claims about saltation in function. We are only talking about genetics. The two work differently.

Your intuition is wrong.

False. We do know how, and that is not the paradox.

I agree. Not one objection you’ve raised, however, is real. Look over the argument carefully.


That’s a valid question to a particular line of inquiry. But it’s not the only question and full elucidation is not necessary to make a solid case for common descent. Basically, we observe genetic drift. Genomes do not remain static. It’s inherent to the the chemistry and biology of life that they drift over time. We’ve measured and calculates the rates of drift. We understand much of the basic mechanisms behind such drift. We also understand that the majority of retained mutations are nearly neutral - basically, non-adaptive. Based on drift and neutral theory, we can see that the genetic variation between humans and chimps, and between mice and rats, fall into the ranges you’d expect from common descent.

The 1975 King and Jukes paper makes a hypothesis for where the morphological or macro-scale changes were likely derived. That’s a very interesting area of work, and it intersect a big portion of the developmental biology field. However, the signal for common descent is strong, even without full knowledge of the developmental changes relating humans with chimps or rats and mice.

(S. Joshua Swamidass) #30

Exactly. It is an important line of inquiry. It is just not the line of inquiry we are taking here.

(Ashwin S) #31

There is nothing surprising about this. The question is what is the causal connection of common descent to mutation rates.

OK, then you must be aware of the major factors that influence mutation rates. What percentage can be connected to the process of replication and hence statistically universal for an organism (for example the error rate in DNA replication due to polymerase slip should be fairly constant). If mutation rates emerge from cellular interactions and environmental variables , it will have nothing to do with common descent in a causal sense… so how can common descent be part of the explanatory framework?

You yourself claim that what you are measuring is not related to function. And termed the lower difference a puzzle in connection to higher functional difference.
That sounds contradictory.

(S. Joshua Swamidass) #32

@Ashwin_s reproductive systems explain the rate of change R, but they don’t explain the starting points. It seems like, given R, chimps and humans had the same starting point genetically about 6 million years ago. That is not explained by R.

(George) #33


The phrase “Common Descent” is not the phrase one associates with the concept of Mutation Rates.

“Common Descent” applies to Speciation through Mutations… regardless of the rate of change.

(Ashwin S) #34

The difference between mouse and rat via a vis human beings is not germane to the above argument. So i am not getting into it.
Yes, this is interesting if we can get repeatable reliable results independent of fossil record. It’s also important to note that we are dealing with extrapolations over millions of years. In such cases repeatability and reliability of the methods is important.Let me suggest a way to test this independent of the fossil record. Take a control organims, say a mouse. Calculate functional relationship T and R between mouse and human beings (I.e f1(T,R)… do the same for mouse and chimps (I.e f2 (T,R))… plot both relationships in a graph and see where they intersect (this should be the CA of humans and chimps). We can repeat this test with different control species and generate n number of dates for when the CA existed. If the results are clustered around 6 million years, and the distant between the two curves using different controls remains similar… it would validate the method…

What were the tests done for repeatability/reliability of the method ? Anything similar to the above…?

(Ashwin S) #35

Correct. That was my basic point.

(George) #36

Okay… so did @swamidass say that it would be?

Maybe even here he was saying Common Descent is unavoidable…and so not the focus of the issue of Mutation rates.

(Ashwin S) #37

I have already agreed that common descent is unavoidable assuming natural causes. Its a direct result of assumptions… I don’t think it can be falsified.

He provided the test to show how common descent explains D=TR being true… since time is an independent factor having to nothing to do with biology… that only leaves R. As far as I understand ,R is connected to cellular level processes during sexual reproduction… and hence has nothing to do with common descent.
Which leaves me wondering, what exactly has been explained.

(George) #38


“Has nothing to do with Common descent” seems like a vast

Reproduction has EVERYTHING to do with the dynamics of common descent… because it is how Common Descent is EXPRESSED.

I you will recall your high school biology… we learn that honey bees are genetically more related to the queen bee than they would be to their own offspring (if they ever had any).

And so because if this traits are passed on to the next generation in a way that maximizes the preservation of the queens SISTERS’ genes than her own.

(Ashwin S) #39

Not necessarily… reproduction by definition is constrained to within a species…
Common descent is all about origin of a species from an older one…
Assuming natural causes, there should be a link… and it’s fair for scientists to do all they can to establish this…
However both of us accept the possibility of miracles. The link could be special creation of some kind for non scientists such as us. After all we are not constraining ourselves to materialism.

(George) #40


With all due respect, I always find it hilarious when someone who rejects evolution tries to tell ME what evolution is.

Common Descent is to Speciation as House Buying is to House Painting (usually).

All speciation involves common descent… but not all common descent leads to Speciation.

So when someone writes a post and says: “There is no proof of common descent”… all I might say is he or she might be worried about being adopted!