“I’m sure there’s a valid transcendental argument out there somewhere, but it doesn’t matter because you already know the Christian God exists, you heathen!”
1 Like
Ewert has a follow-on paper where he introduces a program that is supposed to detect Dependency Graphs patterns in genetic data. That is the only “test” I am aware of anywhere.
I am unqualified to answer the question but will always be weirdly grateful to apologetics (and more specifically, to ID apologists) for helping me deconvert.
1 Like
He also performs a test in his dependency graph paper.
From page 6 of his paper: http://dx.doi.org/10.5048/BIO-C.2018.3.c
Parsimony and FittingMost of the predictions require evaluating whether the data is a better fit to a tree or to a dependency graph. This paper utilizes Bayesian model selection to evaluate which model is a better fit
In the years following Wells’ publication of a hilariously unfounded “hypothesis” about centrioles, he listed something like “use videomicroscopy to study the polar ejection force” on the DI site about his “research.” (It would require use of the Wayback Machine to see if I’m remembering correctly.)
The “hypothesis” was hilariously unfounded but it was straightforwardly testable, and Wells officially had the expertise to do it. My recollection is that he was claiming to be interested in doing those experiments even after scientists had shown that the “hypothesis” couldn’t be right.
So… re the ID movement’s ability or interest in doing science, it’s worse than it looks.
3 Likes
That might explain why, for a while, the internet was rife with would-be apologists making terrible presuppositionalist arguments. They were sometimes called “Sye-clones.”
2 Likes
Not to overstate its merits, but Sye and his clones really don’t do the transcendental argument any justice, nor, for that matter, any favours to the more scholarly theologians and philosophers of religion, needless to say. TAG is not a compelling argument at the best of times, but the childish drivel coming from those bozos is a seriously unflattering reflection, to be fair.
But then Douglas Jones didn’t do much better in the debate I linked above.
“DOI NOT FOUND”
It’s been a while since I looked at that, but I don’t recall that he presented any “test”, only the best “fit” for a dependency graph.
I will look again if the paper reappears.
OR we could just look at the discussion over at TSZ - there was a thorough dissection, IIRC.
It is available on his website above.
No, IDcreationists typically ignore most of the relevant evidence and frequently misrepresent some of the most important evidence.
Please stop. This is not about two different interpretations of the same evidence.
3 Likes
Moreover, he submitted a poster about the hypothesis to an ASCB annual meeting, but wasn’t even there to present it as per custom.
I don’t see how one can be “testing the potential limits of scientific inquiry” while not doing scientific inquiry. Can you explain?
2 Likes
Thanks!
I have to say, that paper looks slick. Another good effort from Ewert, even if I disagree.
@colewd
Ewert is honest about the failing of his model …
Critics will be quick to point out that there are a
variety of mechanisms to explain deviations from the
hierarchical pattern, such as incomplete lineage sorting,
gene flow, horizontal gene transfer, convergent evolution,
and gene resurrection. These mechanisms occur in nature,
but are not included in this model.
… which is good, but admits to some pretty big holes in his model. The comparison is to an incomplete model for evolution, and Dependency Graph (DG) can fill in those holes with an extra node, so of course it’s better here. If we allowed ad-hoc extra nodes in the model for evolution in the same way that is allowed for Dependency Graph (DG) the models should be evilalent. (I can’t test that, but I’ll stand by the statement.)
For evolution we can explain some of those extra nodes, justifying the increased model complexity. Meanwhile DG cannot justify any of those extra nodes, because ID make no predictions. In others words, this DG is making a “Gap” argument for every extra node, arguing the Designer must have done this instead of evolution. The Bayes Factor log-ratios are not so large as to overcome a “reasonable” number of natural mechanisms present in the data set (my educated guess).
If the DG could actually reduce the number of nodes needed compared to an evolutionary tree, and be more parsimonious overall rather than less, then it might start to impress me. As it stands, it appears the Designer is going to a huge amount of time and effort to introduce essentially random noise into the designs, and for no purpose.
4 Likes
Wow, LOL, same for Paul Nelson. I was at 2006 SDB meeting and he had a poster and was a no-show.
Such behavior is hard to reconcile with any sort of sincerity.
I should add that my ASCB report is my own and not hearsay.
2 Likes
Would this be the famed “ontogenetic depth” theory that occasioned Paul Nelson Day?
No, it was the “how to build an animal” trope that is just as shallow and silly:
https://www.sciencedirect.com/science/article/pii/S001216060600546X
Hi Dan
The above reasons are there to explain deviation from the tree model. The deviations from the tree are a natural potential consequence of a separate design or starting points. What evolution cannot do is reconcile how those changes can naturally occur in two populations split from a common ancestor.
The dependency graph infers that unique gene sets are the basic makeup of different animals that are genetically isolated from each other. Since the inference is separate starting points there is no need for a model to reconcile the differences. The current model (population genetics) can then explain how changes occurred in these unique populations.