Mini-journal club on this paper? Testing common descent

Questions arose about testing common descent, in another thread, that I’d like to explore in a more systematic way here. To focus the discussion, I suggest anyone interested read this paper:

Unfortunately, it’s behind a paywall, so you can either contact me for the pdf ( or read the bioRxiv preprint (open access):

Give yourself a week to read and digest the arguments, and then reassemble here for discussion. I’ll kick things off with some questions on February 8.


How about this test? A Test of Common Descent vs. Common Function. I’ve pointed it to you several times. Would you care to comment? It seems to be definitive and quantitative evidence of common descent.

In future, might I suggest providing a full citation? Makes looking for the paper easier. They’re often (in published form) available elsewhere if you just search.

NO DOUBT amongst evolutionary biologists!! how else could it be? Who would be a evolutionary biologist but doubt/deny evolutionism?? i just find it cute the way they put things eh.

When I read the open access link you provided, @pnelson, I found this quote:

“We also find overwhelming evidence that humans share a common ancestor with other primate species.”

Did you realize this is one of its conclusions?

Do you affirm this conclusion?


I’m going to assume that he will argue that the method used is invalid and hopes that others will come to agree with him. What’s your opinion of the method? Because that’s what it really is: a methods paper.

Come on, give Paul a little credit and assume he’s at least read the abstract of a paper before posting it for discussion.

I vote keeping this convo closed to only biologists grad level and up. Then once it’s concluded let everyone else ask questions

1 Like

John is right: it is indeed a methods paper, but it’s not only that. David Baum, the first author, is a really thoughtful biologist (who also works on the origin of life, btw – I’ll post separately a link to a YouTube talk of Baum’s, about his NASA-funded OOL work; aspects of his opening remarks carry echoes of Jim Tour’s critiques, surprisingly). When John said to Ann, in another thread, that all the evidence we have supports the common ancestry of humans and chimps, I immediately thought of this paper, and its “overwhelming evidence” statement.

So the paper provides a good case study in evolutionary inference, with a direct bearing on many of our discussions here.

Many who say that molecular data (in particular, DNA sequence patterns) can be explained only by common ancestry do not realize that the premise (or assumption) of common ancestry actually logically precedes the analysis of the evidence. The assumption of common ancestry in question, however, does not assert that any particular tree of relatedness must be the case. Rather, it’s that some tree (i.e., branching from a common ancestor) unites the taxa in question.

This flowchart, which I made for a Discovery Institute colleague a few years ago, illustrates how most molecular phylogenetic reasoning proceeds:

The David Baum et al. 2016 paper we’ll look at in this mini-journal club, by contrast, asks a question rarely considered by molecular phylogenetics: DO the taxa in question share a common ancestor? In that respect, Baum et al. must consider possibilities hardly ever on the table for evolutionary or comparative biologists, nearly all of whom live and work in a conceptual universe where common ancestry is part of the furniture of reality.

More from me on Friday, Feb. 8…

This is true for any hypothesis being tested. The hypothesis precedes the test. Why is that a problem?


Partial answer:

Tongue-in-cheek answer (sort of):

Austrian stationmaster: Hapsburg Empire trains always run on time.

Passenger: But my train is really late.

Austrian stationmaster: Trains that do not meet their schedules are not “schedule informative.”

Passenger: Say what?

Cryptic, as usual. It seems that DI fellows are unwilling to come out and say what they mean, as a general principle.

Common descent, in most phylogenetic inference, is not being tested. See the flowchart. At no point is the premise or assumption of common descent at risk observationally.

Clear enough?

That’s why the Baum et al. paper is interesting. They claim to test common descent against its competitor (separate ancestry or descent).

No. I don’t agree with your claim, and you can’t just support it by reference to a flow chart. Now of course very seldom is common descent tested explicitly. But one might consider it the implicit null model of any phylogenetic analysis. As with any statistical test, the null model can’t be accepted. At best one could be unable to reject it. In most phylogenetic analyses, we are able to reject it. Now, if you really want to find support for separate creation, you’re going to have to produce a model for what the data ought to look like. Where is that model? Are you, like Baum, willing to accept a set of randomized ancestors with a distribution based on observed site frequencies? Or do you have a different model? Do you have a model at all?

So the problem with my reply wasn’t that it was cryptic.

You just disagreed.

More on Friday.

No. There was more than one problem. If you want to have a serious discussion, you’re going to have to give us more than one-liners.

What happened to this @pnelson?


I still want to do it, but (1) I’m leaving shortly with my wife for 10 days in the Galapagos Islands on a National Geographic ship, (2) have to finish my taxes (groan) before then, and (3) I’m preparing to lead an online research seminar next week on “the Function Wars.” Just FYI, here’s the abstract for my seminar:


Following Project ENCODE, and the widely-announced “demise of junk DNA,” a theoretical and experimental ‘Junk DNA Counter-Reformation’ was launched by senior scientists such as W. Ford Doolittle, Dan Graur, Sean Eddy, and many others. Junk DNA was alive and well, they argued, because the criteria for function employed by ENCODE researchers were far too non-specific and unsupported by detectable phenotypes. Moreover, theoretical arguments (e.g., genetic load) strongly ruled out functionality for most of the human genome, not to mention other genomes. Under stricter criteria, such as the “selected effect” concept for function, junk was inevitable. In this presentation, Paul Nelson will sort out the issues, and show how a nuanced design approach gives the best prospects for genetic discovery and for dealing with apparent suboptimality or non-function. Plenty of good biology, with only a Minimum Daily Requirement of philosophy and logic.

First off: nice. We recently had the Grants give a talk at my university about their exploits in the Galapagos, and it looks like an amazing place.

Second, I don’t suppose you could give us a couple of sentences summing up the conclusion you have planned for the seminar? My guess would be “the vast majority of the genome is probably functional”, but I don’t know if you have a less strong opinion on this than most YECs.

Regarding this thread in particular, no worries, it can wait.

1 Like