I’d go so far as to say that it’s frequently true. Then again, it’s frequently false. It’s probably not the best thing to try to convince Bill of. Then again, again, could there even be a best thing?
2 Likes
I can appreciate how painfully embarrassing it must be for you to have your ignorance and deception exposed in such a public manner.
But if you want to get back at John and Michael, there’s a really easy way you can do it: Just post a series of Howe-type diagrams, some from species that actually exist, and some that are completely made up. Then ask them to tell which are which. Since it is highly unlikely that a random diagram would produce a nested hierarchy, they should be able to do so. But only if they are correct in their claims that a nested hierarchy is indication of common ancestry and that all extant organisms are related by common descent.
So are you game? Here’s the perfect opportunity for you to save face.
3 Likes
Well one problem is you can of course just make up a nested hierarchy with a tiny bit of effort, so Bill could decide to just force one on his data for no reason in particular. The problem is there’s no reason why independent creation of organisms should happen to produce one as a byproduct of the process of independent creation unless you’re explicitly using a branching genealogical process to create. But then it’s not actually independent creation.
1 Like
If he did that, it would be an admission that he is wrong. i.e. it would show that the nested hierarchy cannot be expected to arise from separate creation unless it was deliberately created.
1 Like
I think Winston Ewert showed that this claim is suspect.
From his paper.
The JavaScript applications fit the tree or the depen- dency graph better than the null model. However, the dependency graph is preferred to the tree. This again confirms one of the predictions, software can exhibit a hierarchical signal while being produced by a dependency graph. Nevertheless, it still fits the dependency graph better than the hierarchical pattern.
Doesn’t work for DNA sequences.
2 Likes
Can you summarize this “dependency graph” position?
1 Like
No. He can only quote, since he doesn’t actually understand it any more than he understands nested hierarchy.
1 Like
Um that’s not the answer I was expecting. I asked Rum to explain what it is. Maybe you can answer that.
Sorry, I thought you were asking Bill. The easiest thing to do is point you toward Ewert’s paper, if you haven’t seen it. I don’t understand it completely myself, and the methods in particular seem very incompletely described, but I can see a number of fatal flaws. It’s been discussed at length here.
1 Like
Short summary: The claim is that dependencies between pieces of code in different software programs(such as java script applications) will produce nesting hierarchical patterns that is tree-like, but actually fit a so-called dependency-graph better than a strict tree. These pieces of code can be considered “modules” of the software that depend on each other for their function.
The claim is that we are seeing a similar thing with genes in organisms constituting “modules” that depend on each other to function(so the reason we see a particular set of genes in the mouse or the chicken is that these genes depend on each other, as in that pattern is necessary to get a mouse, or a chicken), and therefore the hierarchical pattern in shared genes among species are best explained as a byproduct of the independent creation of these organisms using a dependency relationship among their genes. So in other words, these patterns in the pieces of code found in certain software programs is supposed to provide a counterexample to the inference that only a branching genealogical process produces nesting hierarchical patterns in the data (it’s not clear to me whether there truly is no branching genealogy involved in the putative “independent creation” of these software programs).
The claim is, further, that the hierarchical pattern in the data of shared genes actually found in living organisms fit a dependency graph better than a tree.
The dependency graph doesn’t deal with DNA sequences, it only considers orthologous genes as “modules” with co-dependencies, but does not provide an explanation for the hierarchical pattern in the DNA sequences themselves. It doesn’t deal with things like the fossil record, consilience of independent phylogenies derived from other pieces or data, physiological data/morphology, or anything like that.
And yes, all this has been discussed on this website already, with Winston Ewert himself, who agreed his work had several shortcomings.
2 Likes
Link to the thread?
This is perhaps the biggest problem, as it assumes the co-dependencies purely on the basis of the observed distribution of gene families (not orthologous genes) among taxa, and it constructs purely ad hoc “modules” for each separate distribution pattern.
1 Like
Anyone who argues that software and biology share hierarchical principles understands neither or misrepresents both.
In software, common modifications to the tips of trees is routine, as new technologies are grafted into separate existing programming languages and platforms. Outside the limited scope of HGT, Biology cannot do this, precisely because it depends on descent and not design.
4 Likes
Hi Ron
I agree to some degree. The original version of software can certainly have dependent modules that are selected. Upgraded new versions will add to that design as you are claiming.
Biology will copy itself with some slight modification through recombination and mutation. What we are observing in some cases however is dramatic genetic modification including new genes in many cases. How do you account for this?
You have been told this so many times. There are mechanisms that give rise to new genes and end existing ones. Duplication, fusion, etcetera can facilitate the emergence of novel genes.
Bill you asked for multicellular eukaryotes. I gave you some in the Venn diagram below, so tell me, where they independently created or not?
Stop ignoring me.
Edit: I mixed up identities here. This Venn diagram is for unicellular eukaryotes. The one below is for the multicellular eukaryotes:
2 Likes
How do you know those mechanisms account for the pattern? What supporting evidence do you have?
Bill stop sidetracking. Look at that Venn diagram with five organisms, all multicellular eukaryotes as you wished for. Were they independently made or not?
2 Likes
The answer is I don’t know.
You claimed to have a mechanistic explanation for the gene pattern. If you can defend that claim the game you are playing with Venn diagrams does not matter.
Again what evidence do you have that the mechanisms you cited created the pattern you are observing?
So how did you arrive at the conclusion that Howe’s diagram (made exactly the same way as this one) reflects separation creation events?
4 Likes