Bayesian inference and the teleological argument

I don’t think you know what “model” means in a scientific context.

Population genetics is not a model.


OR they are due to natural processes which can cause this appearance.

Ridiculous. The reconciliation you say is needed for natural processes (your usual nonsense) should be equally needed for any other valid hypothesis. You are making a Special Pleading.

As it stands, I find Ewert’s evidence entirely underwhelming. Designers tend to reuse parts because it is efficient, not to add weird exceptions. Again, if a DG showed better parsimony that could be good evidence, but that’s not what Ewert has shown.

AND what he has actually shown, IMO, is there are gaps in genetic data. We already knew this.


Hi Dan
The need for reconciliation of how the differences occured is based on the claim of a reproductive connection between species. Winston does not make this inference in his DG model. Again, if you do not claim reproductive connection then you simply can model and test changes in specific populations.

This is true and the model is indeed complex. The simplest way to express gene differences and similarities appear to be the Venn diagrams.

What we did not know is that a model that is used in software design would be a better fit to the gene data than a tree.

Which is your usual claim, demanding details of genetic changes for each and every generation separating any two species.

Winston does not make this inference in his DG model. Again, if you do not claim reproductive connection then you simply can model and test changes in specific populations.

In that case, he absolutely make this claim, because he starts with the existing tree.

(In fairness, I doubt that Winston understands your “reconciliation of how the differences occurred” thing any more than people here. I can’t blame him for a non-existent problems he did not create.)

Thank you Bill for admitting that there is no model, and thus no explanation. Therefore “separate starting points” cannot be an ‘inference to the best explanation’. I’m glad to see this resolved.

I would however like to question the implicit equation of the ‘Dependency Graph’ with “separate starting points”. Does a positive outcome for DG actually demonstrate separate starting points, or simply the case that we do not know what the true relationship is? Conflating the two would appear to be blatant ‘God of the Gaps’ argumentation. If a detailed positive model/explanation cannot be articulated for “separate starting points”, then it would appear that a testable hypothesis cannot be created for this scenario. The scenario would simply be too lacking in specifics (i.e. vacuous) to be testable.

Beyond this, I would also question the degree to which Ewert’s methodology has been validated. To date, the only acceptance it seems to have achieved has been from within the ID community. But, as I and others have demonstrated above, that community lacks credibility. So I feel some external validation is needed for his methodology to win any acceptance.


One hint about this can be found in the citations of Ewert’s paper. The data at Dimensions show 10 citations in 6 years. That alone indicates essentially zero interest in the paper, let alone validating (or using) the findings. It gets worse (or funnier?) when you look at the citations, which you can see in the attachment.
Ewert paper citations.xlsx - Sheet1.pdf (36.6 KB)

Of the ten:

  • Six are obscure papers about forensic mental health or aspects of sexual violence. I have no idea why they are citing Ewert’s paper. :thinking:
  • Of the remaining four, two are about philosophy or engineering (the latter is a book chapter and not a peer reviewed paper), and the other two are by Cornelius Hunter. :laughing:

How about starting with any two species with shared and different genes such as rats and mice?

His claim is that the DG is supporting common design. From his paper:

The dependency graph hypothesis draws on the idea of common design,
by having reusable modules, as well as functional requirements, by restricting the reuse of modules via the dependency graph. The concept of a dependency graph draws not from an ad-hoc attempt to explain the data, but the actual process used to develop software. It is based on behaviors and practices that intelligent agents are known to use, not simply processes necessary to explain the data.

How about starting with any two individuals with shared and different genes, such as a parent (or parent tuple, how ever many it takes to produce offspring in the species at hand) and one of its offspring? How do you test for reproductive relatedness at all, in light of the fact that less than every DNA segment in an organism is an exact copy of the sequence at the corresponding locus in their parents’ genomes? Indeed, how do we define what a ‘corresponding locus’ even is, if less than all of an organism’s genome is a recombination of parent genomes?


But his dependency graphs aren’t actually dependency graphs. For that, he has to show that his modules are actual functional modules and that there are real dependencies rather than chance associations. He provides only a single example of a module, for echolocation, and it’s entirely hypothetical. He would also have to explain why the sequences of the genes in these supposed modules are neither identical nor independently different but instead show nested hierarchy.


Except that Mus (“typical mice”) and Rattus (rats) are genera not species. And the larger the populations (genera over species, species over individuals), the larger the data sets needing to be documented and analysed. @Gisteron’s suggestion above of analysing a parent(s)/child grouping would seem more practical, at least as a starting point.


Credit for italicizing the genera, but you also need to capitalize them.

Corrected. :slight_smile:

YOUR claim, is that ToE “need(s) for reconciliation of how the differences occured”. The DG assumes the evolutionary tree as a starting point, and therefore has the same need.

You bring this “need” up often, but never explain or define what it means.

If you will first define the question to be answered, maybe I’ll take a poke at it. What about that, exactly, needs to be explained?

To help you get started, here is a series of statements. Let me know which ones you disagree with.

  1. All species contain a degree of genetic variations; most genes are the same, but some are different.
  2. The above should also be true of the LCA of rats and mice.
  3. Rats and mice have some genes in common, and some that differ.
  4. The degree of genetic variation between rats and mice is (likely to be) larger than the genetic variation with the LCA.

OK, go.


The tree itself is a pattern. It alone does not establish the cause of the pattern. Again, Winston does not claim reproductive connection as the common descent claim does.

What needs to be established is a reproductive connection between rats and mice. The obstacle is reproductive variation beyond what we observe in species where we can more easily infer common descent as between generations of mice and generations of rats.

Not sure there is functional gene variation in like species if this is what you mean. There maybe slight sequence variation in the same genes. ie blue eyes vs brown eyes.

Disagree as the LCA population should share the same genes with genetic variation that can be reconciled using population genetic models.


Agree but not relevant to the discussion as we are discussing gene differences which is more than genetic variation.

But how do you infer descent between generations of mice or generations of rats? How do you actually conclude that two individuals are related, outside of literally observing the birth or believing the words of someone else claiming it?

How do sequences not present in parents come to be in children?


BUT he does allow for common descent, therefore the same problem, whatever you think that is, must apply.

OR you can continue with your Special Pleading for Design.

For which I refer you to @Gisteron’s previous comment:

AND his following comment too :point_down:


Of course there is functional variation in species. For example, some humans cannot digest lactose as adults, but others can. This is both functional and adaptive.

It is absolutely relevant. Gene difference exist within every species. Closely related species have fewer differences than more distantly related species. Variation naturally increases over time (genetic drift). In situations where, because of these differences, one sub-population can no longer produce offspring with another sub-population, we begin to see the emergence of separate species.
Granted that line between species is a bit fuzzy, but I think that definition will serve for this discussion.

Do you see any obstacle to genetic drift?

I think your objection/obstacle has been answered. If you disagree, then please state your reasoning for why new species cannot emerge.


It is May 17th, 2024. Bill has not been able to answer such questions since what, 2015? There isn’t a future in which this will change or where his capacity to engage meaningfully substantially departs from that of a rock. I’m not fully persuaded he isn’t, in fact, a mineral of some sort.


The description I was remembering is at the Biologic Institute website under “People”. The entry for Wells appeared in October 2009 (yay for the Wayback Machine) and has been unchanged since (!!):

Jonathan Wells
is a cell and developmental biologist with a PhD in Molecular and Cell Biology from UC Berkeley. His current research includes experimental testing of a hypothesis about centriole function with implications for cancer, and theoretical work on the role of endogenous electric fields in establishing spacial coordinate systems to control morphogenesis in animal embryos. His work has appeared in BioSystems, Development, and Proceedings of the National Academy of Sciences.


@colewd and Andrew:
Cc: @Dan_Eastwood

Please note that this thread is yet another iteration of the hopeless and endless dispute “can you prove the existance of God with a mathematical equation.”


Isnt it ironic that on a web-page - -designed and created by a Christian in order to ease communication with the Evangelical wing of Christianity - - most of our bandwidth seems to be devoted to telling Creationists that the Bible is not only wrong about Adam & Eve, but that God had nothing to do with creation.

If i were to poll the non-theists at PeacefulScience, how many of you would confess to @swamidass that you have virtually no interest in explaining to fellow non-theists how the GAE reconciles science with Adam and Eve?

No such confession is necessary. Josh knows that GAE isn’t aimed at non-theists, and doesn’t expect non-theists to explain it to other non-theists.