Earlier this year, a popular evolution YouTuber, Gutsick Gibbon, or Erika, created a video response to my post here at Evolution News, “Do Statistics Prove Common Ancestry?” I had reviewed a paper by Baum et al. (2016), “Statistical evidence for common ancestry: Application to primates,” and how it presents a flawed and weak argument for separate ancestry that ignores the possibility of common design.
Erika is currently pursuing her Master’s of Research in Primate Biology, Behavior and Conservation and is the creator of hundreds of punchy, entertaining YouTube videos. Her channel’s primary focus seems to be debunking Darwin-skeptics. Unfortunately, she does not seem to apply an equally critical eye to evolutionary theory.
An Initial Disclaimer
Before going further I want to remind you that intelligent design (ID) is compatible with both common ancestry and non-common ancestry views. Some of my colleagues here at Discovery Institute support common ancestry while others (like myself) are more skeptical. That’s OK! We all agree that there is evidence for design in nature. Some of us skeptics are interested in exploring potential models where ID and non-common ancestry histories of life intersect. Design does not rise or fall with these models, but they are interesting questions to explore.
Critiquing the Critique
Erika’s first critique can be summarized as a complete misunderstanding of ID proponents’ objection to the paper. We will deal with that in a post tomorrow.
Her second critique is that ID proponents shouldn’t expect others to test their models, but should test the models themselves. Anyone is welcome to test ID concepts if they like, but I don’t think that ID proponents were expecting Baum et al. (2016) to test the hypothesis of separate ancestry. Rather, the paper carried out the normal scientific process where one group of scientists tests another group’s scientific hypothesis independently. Only, in this case they tested a hypothesis no one supports — more on that later. Perhaps most important, ID proponents are involved in testing models of separate ancestry and the example here was provided in the original post.
Her third critique responded to my key point — there are two known mechanisms (design and ancestry) that can produce genetic similarity. Therefore, genetic similarity should not always be used to provide exclusive support for ancestral relatedness when other explanations are possible.
To elaborate on her third critique, Erika argues that there is no genetic demarcation or separation that would mark a stopping place for comparison between species and higher orders of phyla. She is clearly ignoring reproductive barriers here. While I don’t think this argument addresses my bolded point above, I am quite curious what she imagines this stopping point would look like if in fact separate ancestry were correct? I speculate that in such cases people expect the only evidence for a discontinuity in biological relatedness would be a vastly different genetic code for each organism or species. This seems to me a false expectation, because human technology shows that even separately designed structures can have deep similarities that go down to their very blueprints or encoding information. Given that, a design hypothesis would lead us to expect functional similarities. I would also say that there are reasons that a good design would make use of a highly similar genetic code for all organisms.
In this part of her argument, Erika also discusses “Last Thursdayism.” She says because a seemingly hierarchical ancestral pattern exists, if separate ancestry were correct, the designer must be deceptive to leave us such a pattern. In case you aren’t familiar with Last Thursdayism, it is a concept that a creator or God could make things look a certain way (billions of years old for example) even if he had created everything last Thursday. While I agree there are problems with Last Thursdayism, Last Thursdayism isn’t relevant in this case. There are straightforward reasons to expect some degree of tree-like patterns even in a non-common-ancestry-related dataset.
If the seemingly “deceptive pattern” exists for a functional reason and has a “good design” explanation, then there really isn’t a “deceptive pattern.” The “deceptive pattern” is imposed only by materialist lenses and a poor understanding of functional reasons for the similarities.
To summarize the problems with her third critique, as emphasized in my original post, we know and observe two mechanisms that can result in genetic similarity. Design is one (think genetic engineering) and ancestry (think reproduction) the other. Because two known mechanisms exist to produce genetic similarity, that means, in and of itself, that genetic similarity does not provide evidence for ancestral relatedness. Certain patterns of genetic similarity may do this, but a design pattern, which isn’t randomness, was not considered in the Baum et al. (2016) paper and isn’t being considered by many in the academic community. That’s what ID proponents are trying to change.
Erika’s fourth critique is that Winston Ewert’s dependency graph (Ewert 2018) is not an actual model of separate ancestry. Winston’s central thesis is that the nested hierarchical pattern observed in subsets of genes is better accounted for by a dependency graph. Erika acknowledges this is outside of her field, but she quotes Joshua Swamidass to dismiss it as a model. I’ll talk more about her specific points in a later post.
Finally, Erika’s last point is to address my argument that Baum et al. (2016) “cherry picked” which genes they would use when constructing their phylogeny: they only used genes they claimed were “phylogenetically informative,” which could imply a stacked deck. She really did not address my argument and instead made a comment about orphan genes.
I did not feel that Erika provided evidence for how (experimental) or why (conceptual) common design could not result in genetic similarities between species. Instead, there is evidence of design-dependent genetic similarity exploding all around me. I see it in the artificial selection for dogs (breeding for specific traits). I performed it myself in the lab using recombinant DNA technology. And I see it being dreamed about for the future as bringing about incredible advances in human health using CRISPR-Cas9. These are all proof-of-principle examples that design can and does produce genetic similarity in different organisms. Because this mechanism is well established, when we observe genetic similarity, we can’t refuse to include design in the conversation.
In my next post I will explain why I think Gutsick was confused about the objection I raised previously to the separate ancestry model in the Baum et al. (2016) paper and attempt to explain the ID position more clearly.