Is homoplasy hidden from non-specialists?

By Richard B. Hoppe

This started as a comment on my earlier post responding to a comment by John Harshman, but it outgrew comment length so I’ll do it as a post. It may appear to be beating a long-dead horse that’s suffered enough, but there’s an aspect of Gauger’s commentary that is again part of the Disco ‘Tute’s efforts to undermine common descent and impugn the credibility and honesty of evolutionary scientists in general that deserves attention.

John Harshman wrote

I don’t deny that the green screen is a valid target for ridicule. And it is indeed a fine metaphor for the whole DI exercise (though I really like “cargo cult science”). I had two points:

  1. There’s been too much attention to the green screen in proportion to its importance. This may be because it’s a subject for which those who don’t know much about the biology feel free to contribute. It might be that the green screen is equally understandable and meaningful for the ignorant public, and so should be emphasized, but I don’t think that argument is a strong one.

Nor do I think it’s particularly strong except as a manifestation of the ‘business as usual’ approach to rhetoric of the Disco ‘Tute. And having earned a degree in anthropology many decades ago, I too like the “cargo cult science” characterization.

John went on

  1. Many of the posters on the subject have made untrue conjectures, particularly the notion that they didn’t pay for the photo or that they don’t have an actual lab they could have used. That’s where the Jesuit triumphantly produces the live dog.

Yup, and that’s a valid criticism of some of the comments on these posts. John continued

It isn’t just in SINEs that there’s very little homoplasy in hominid evolution. The proportion of homoplasy in simple SNPs is low enough that ignoring it entirely still gets you the correct tree. And this is true even if you use fast-evolving sequences like mtDNA. Lineage sorting is a bigger problem, though only for the African ape trichotomy, and you can ignore that too if you concatenate as few as 5 or 6 genes.

Apropos of the scientific issue, here’s a discussion, aimed at non-specialists, of an approach to deriving a phylogeny (in a paper by John Harshman, no less) that uses multiple loci to mitigate the homoplasy issue:

Instead of relying on just one or a few regions of nuclear or mitochondrial DNA (mtDNA), this analysis relied upon using a phylogenomic approach, which analyzes data from many regions within the avian genome. In this case, the research team analyzed 20 loci that are widely dispersed throughout the genome, comprising sequences from both protein-coding (30 percent) and non-coding (70 percent) regions.

And wouldn’t you know it, that discussion is on a widely read blog aimed at lay people. From the Abstract of the paper:

Phenomena that can mislead phylogenetic analyses, including long branch attraction, base compositional bias, discordance between gene trees and species trees, and sequence alignment errors, have been eliminated as explanations for this result.

No, Ann, real professionals don’t ignore potential problems with phylogenetic reconstructions. They directly address them.

In her Facebook response to various criticisms of her video, Gauger wrote

  1. About homoplasy being a hidden secret: it’s hidden from non-specialists. The technical literature is aware and trying to deal with it. Just see the post Confusing similarity for a discussion of two mainstream articles 12 years apart. But you would not know this from listening to Dawkins or any other evolutionary evangelist.

Hm. Convergent evolution is very nearly a synonym for homoplasy. In her Facebook response Gauger defines homoplasy as “2. About homoplasy: it means similarity of a trait or genetic sequence not due to common descent.” That’s convergent evolution, and Dawkins has a whole chapter on convergent evolution in The Ancestor’s Tale. The book description on Amazon.com says

Dawkins’s brilliant, inventive approach allows us to view the connections between ourselves and all other life in a bracingly novel way. It also lets him shed bright new light on the most compelling aspects of evolutionary history and theory: sexual selection, speciation, convergent evolution, extinction, genetics, plate tectonics, geographical dispersal, and more. The Ancestor’s Tale is at once a far-reaching survey of the latest, best thinking on biology and a fascinating history of life on Earth. (Emphasis added)

I won’t mention, say, Simon Conway Morris, who is fixated on convergent evolution as support for his claim that the evolution of critters with human-like intelligence is nigh unto inevitable. Conway Morris doesn’t reject common ancestry on that account.

So Gauger’s claim about the issue being hidden from non-specialists is just flatly false. It’s right out there in plain site in books by professional scientists that are specifically aimed at non-specialists. Gauger not only misrepresents the science, she misrepresents the information directed at and readily available to non-specialists. But then, what else would we expect?