Evolution may occur through a dependency graph, not the conventional simple tree | Hacker News Reader

Evolution may occur through a dependency graph, not the conventional simple tree

evolutionnews.org

16 pointscetico14 comments

This is an article from an antievolution, pro-intelligent design, 'science news' website, run by the Discovery Institute. It is promoting a paper from their own journal.

It was not peer-reviewed. And was created and funded based on ideological presuppositions.

Meanwhile in the real-world, it has long been known that the evolutionary tree is a graph. But the degree of lateral gene transfer is typically thought to be small. Even more so for more complex taxa. (Perhaps more common in bacteria, for example).

Please don't confuse this website or paper with science. The author would need to do a lot more work in a very different intellectual context to make it so.

Thanks for this reply. Since most people (myself included) are not familiar with the state of the art, this article sounds somewhat compelling, given it describes a plausible correction to the prolific idea of evolutionary trees. At least it's immediately clear they are overstating the impact of this work though, given the presentation style and inconsistency. Claiming the graph model and simultaneously dismissing LGT with no further explanation is nonsensical.

Indeed, thanks for clarifying. I apologise for posting pseudo-science in the first place.

(Although a little bit of divergent thinking is healthy for the mind)

Initially looking at the article, it sounded intelligent, talking about measuring the fit of data and a theory. But yup, another article mentions the soul and talks about human exceptionalism.

Still much more advanced than the average creationist who only offers doubt and a supernatural explanation

It nicely deferred the giveaway phrase 'intelligent design' until very late. But they call evolutionary biologists 'evolutionists', and find that this new model refutes the whale fossil series (creationists really hate the whale fossil series). And, as always, it is a model without a mechanism. Because they already know the mechanism. God did it.

Their about page (https://evolutionnews.org/about/) lists "Discovery Institute" as the copyright owner.

From wikipedia on the Discovery Institute: "Discovery Institute is a politically conservative non-profit think tank based in Seattle, Washington, that advocates the pseudoscientific concept of intelligent design."

Would this imply that species have a clever way to acquire "modules" from other species?

Would that happen via crossbreeding? Or some mysterious form of gene "absorption"?

Although the author's opinion appears to be intelligent design:

"As a consequence, all purported evolutionary trees and sequences become highly suspect, including such icons as the whale and human series. For they are based on similarities of traits between species, and similarities are an unreliable indicator of common ancestry as implied by the trees’ typically low adjusted consistency indices. Instead, similarities appear to be the result of a designer reusing design modules in different species to meet common goals.

Ewert’s article represents only the first step in evaluating and developing his framework. Still, the significance of this research cannot be overstated. The dependency graph model explains why subsets of the biological data crudely fit a tree pattern and why so much of the data is incongruent. It also makes clear predictions on the results of future studies on the distribution across species of both physical traits and similarities in molecular data. Finally, it should lead to a robust and innovative research program based on the intelligent design framework. "

https://en.m.wikipedia.org/wiki/Horizontal_gene_transfer

The OP article specifically writes

> These disappointing results have required evolutionists to devise several ad hoc mechanisms to explain the ubiquitous inconsistencies. Examples include lateral gene transfer (LGT), differential gene loss, and convergent evolution. Yet, the widescale appeal to LGT has been seriously questioned.

LGT is a synonym for HGT. I'm not sure why the article brings this up, since the model (DAG vs tree) and the mechanical implementation (LGT, etc.) are different things. The need for a mechanical explanation (defined by observation) does not change with a new model (also defined by observation). That said, new models can provide the right mental states to speed up discovery of correct mechanisms, so this work may have some significance, but it will be impossible to tell a priori.

The "conventional simple tree" is just a depiction of data put into visual representation easy for humans to understand.

On the lineage of species, the actual heritage and relationship through evolution is more of an acyclic graph. We just have species that we know fill the gap between other species enough to know that a tree is the simplest way to depict evolution over billions of years, just like a tree may depict a family genealogy or family tree but even then we know its not all linear because humans have married between families and been incestuous at times.

May be they are Modules with feature toggles and evolution just turns on/off what is there already.

This is part of the post-intelligent-design attack on evolution. The idea that we couldn't have evolved, because evolution can't create information. It merely turns on and off information that is already there. The website linked here is an intelligent design site. But the idea in your comment would be used in various forms on a number of other creationist sites. It allows them to imply that these feature modules were created by God. And evolutionary change is merely flicking switches.

Yeah, I noticed as I made my comment that this was a likely conclusion. Either God or Aliens. Or Both.

However, I didn't know that the web site itself was an intelligent design site.

Reply on news.ycombinator.com