Darwinian time trees don’t really work

_{Denyse O'Leary

January 12, 2021

Darwinism, Intelligent Design, Tree of life}

Share: Facebook; Twitter; LinkedIn; Flipboard; Print; Email

They are “deeply flawed,” say two researchers:

Two theorists have caused a stir in evolutionary circles, claiming to have proven that Darwinian phylogeny efforts (tree-building) cannot be constrained to one “best” answer. In fact, any proposed tree is no better than an infinity of other trees. They can’t see the tree for the forest. p1 The theorists are Stilianos Louca, biologist at the University of Oregon, and Matthew W. Pennell, evolutionary biologist at the University of British Columbia. Their paper that started the controversy was published in Nature, “Extant timetrees are consistent with a myriad of diversification histories.” A timetree is a phylogenetic tree supposedly calibrated by the appearance and disappearance of organisms. An extant timetree is a timetree calibrated using living organisms. The news from the University of Oregon, “Researchers find flaws in how scientists build trees of life,” sums up the paper’s thesis that “long-used approaches for reconstructing evolutionary paths are deeply flawed.”
Evolution News, “Controversy Arising: Timetrees Unconstrained” at Evolution News and Science Today

The paper is paywalled.

We hope those people’s careers are safe.

The controversy arising about timetrees is something to think about when hearing confident-sounding presentations about the history and evolution of life. When scientists speak glibly about adaptive radiations, early bursts of diversification, global extinctions and all the rest, what do they really know? They weren’t there. They take bits of bone, molecules from eye of newt and bat wing, and conjure up fantastical scenarios of an evolving world of universal common ancestry driven onward and upward by natural selection alone. But if the model is just one of an infinite number of congruent timetrees held together by unrealistic adjustments, the world picture may never have existed except in the crystal ball of the imagination.
Evolution News, “Controversy Arising: Timetrees Unconstrained” at Evolution News and Science Today

Well, sure, but is it safe to say that?

The article goes on to talk about refutations but so far it seems like a civilized discussion. That’s evidence that Darwinism is losing its chokehold on thought.

Comments

You made the comment on March 22nd, clearly referring to the whole world (and comparing tothe global impact of flu: https://uncommondescent.com/intelligent-design/pass-me-a-corona/#comment-695811 Indeed, on march second there where approx. 13k deaths globally (https://ourworldindata.org/grapher/total-covid-deaths-region?time=2020-01-11..latest) hence your logic that deaths would double only one more time. It actually took less than a week to double. The reported total has actuallyl doubled 7 times, and thatis likely an undersetimatae. It was not hard for people who knew anything about infectious diseases to see this impact of the pandemic coming.orthomyxo_{January 22, 2021
January
01
Jan
22
22
2021
05:20 PM
5
05
20
PM
PDT}

Orthomyxo: Please get your facts correct. I've already corrected you once. The estimate of 27,000 was based on the Chinese experience, and was for the UNITED STATES only, not the world. Why would I predict 27,000 deaths for the entire world when we knew that 30,000 has already died in China. What logic were you using. Just for the record, the other day it came out that China lied about their numbers by 10-fold. Thus, the corrected figure, based on the correct China experience would be 270,000. The current figure for the US is wildly skewed. The "actual" Covid deaths in the US thus far and the this corrected figure are right now not too far off. This will of course change. However, you must admit that it's hard to generate good estimates when you're being lied to. Does any of this matter at all to you?PaV_{January 22, 2021
January
01
Jan
22
22
2021
04:50 PM
4
04
50
PM
PDT}

What a cast of characters, ET things vitD could have prevented the pandemic, PaV is unchastended after his prediction 27k ppl would die in the entire world has shown to be out by orders of magnitude and Es is so tied good conspiracy theory he thinks the Georgia bureau of investigation is in on the Joe Biden's election fraud scheme. Don't really mark yourselves out as being people to be taken seriouslyorthomyxo_{January 22, 2021
January
01
Jan
22
22
2021
03:14 PM
3
03
14
PM
PDT}

I see there are no limits to your ineptitude, Ortho.ET_{January 22, 2021
January
01
Jan
22
22
2021
05:57 AM
5
05
57
AM
PDT}

Orthomyxo You told me on the other thread that the audit was conducted by the Georgia bureau of investigation. Thanks for making *my* point!es58_{January 21, 2021
January
01
Jan
21
21
2021
10:05 PM
10
10
05
PM
PDT}

Orthomyxo: Were the experts right when they said 2,000,000 Americans would die? Were they right when they revised this and said only 65,000 would die? Were the experts right when they further revised their numbers and said 165,000 Americans would die? All these numbers were come by using models. My argument then was that there was no reason to "lockdown" the economy as a means of coping with this virus, and I pointed out severe historical instances of the flu as reasons not to. I later changed my mind and said that the "lockdown" was justified only from the standpoint of keeping our hospital and medical services from being overwhelmed, but not as an effective strategy against the spread of the virus. Evidence now suggests that "lockdowns" don't work. In my home-state of California, I've witnessed this first-hand. Eight weeks ago, we went under one of the most restricted lockdowns in this country, and our numbers didn't slow down or level off; rather, they exploded. As to the virus, most people who get the Covid virus experience mild flu-like symptoms. There is a certain subset who, after six or seven days, experience a severe respiratory illness which does prove fatal in those 65 years and older. So, no, this isn't just something like the regular flu. What I didn't appreciate sufficiently was the contagiousness of the virus, which we struggle with to this day. However, that said, Dr. Fauci has said that it's his belief that this virus will eventually become like a cold flu virus, coming back each year, as other corona viruses have done, and with our population having built up immunity to it. Again, my problem was with the lockdown, the wisdom of the lockdown. And as evidence gathers, I think the decision to lockdown will one day be viewed as one of the most foolish decisions taken by the medical community, really by Trump. Getting back onto the present issue, I don't consider myself well-informed on these matters and have sensed from the beginning that you are. So, there's no need for me to pass myself off as an expert. Yet, your statements about the ENV article have been declarative and not illuminative. If you want me to acknowledge that ENV got it "wrong," then the best way to do that is to lay out an argument. If you want to know why I keep making these "comments," it's because I'm trying to understand the point you want to make. Here's a nicely laid out argument on this topic. I would recommend it to anyone who is interested in this discussion. From what I've read, I think that on-balance the ENV article didn't sufficiently clarify the phylogenetic approach from the paleontological approach (spoken about in the above cited blog article), and to that extent, their criticisms miss the mark. I believe that was point, more or less, you were making.PaV_{January 21, 2021
January
01
Jan
21
21
2021
05:16 PM
5
05
16
PM
PDT}

I see there are no limits to your wrongness, ET.orthomyxo_{January 21, 2021
January
01
Jan
21
21
2021
12:47 PM
12
12
47
PM
PDT}

Take away the flu vaccine and covid and the flu would be the same. And the best part is they can both be easily fought off with OTC supplements.ET_{January 21, 2021
January
01
Jan
21
21
2021
12:21 PM
12
12
21
PM
PDT}

Ortho- you aren't even responding to what I am posting.ET_{January 21, 2021
January
01
Jan
21
21
2021
12:20 PM
12
12
20
PM
PDT}

PaV, you clearly don't know anything about this topic, so why do you persist in making these comments. I've said it all as clearly as a I can. This issue is with models that are built on top of trees. The ENV story is written as if the issue is with the tree-building itself (and the News article here compounds it). It's not true that this paper shows time trees are unreliable without fossils (because the paper is not about the reliability of time trees). Let me ask you another question. Last time I intereacted with you on this site it was because your own critical abilities had led you to belive the covid-19 pandemic was very little threat and unlikely to amount to more than a regular flu season in the US. Has you experience of being so horrifically wrong about this let to any adjustment in the way to you assess evidence and consider the opinions of experts?orthomyxo_{January 21, 2021
January
01
Jan
21
21
2021
11:56 AM
11
11
56
AM
PDT}

Orthomyxo: Aha! Identifiability! From Morlon, Hartig and Robin paper, their first paragraph:
Louca & Pennell2 consider the homogeneous (i.e. lineage-independent) stochastic birth-death process of cladogenesis traditionally used in macroevolution to test hypotheses about how and why rates of speciation (birth, [lambda]) and extinction (death, [mu]) have changed over time t. They show that for any given time-dependent speciation function [lambda] and extinction function [lambda]>0 or [mu]> 0, there exists an infinite set of alternative functions and [lambda] such that the probability distribution of extant trees under the corresponding birth-death processes M and M* is identical. Consequently, M or M* yield identical likelihood values for any given empirical tree. This identifiability issue is certainly both interesting and unfortunate, but what are its implications for phylogenetic-based diversification analyses?
And, later on:
By considering all possible diversification functions, Louca & Pennell implicitly subscribe to a different method of scientific discovery, where the goal is to learn [lambda]and ? from the data alone. After finding that these quantities are not simultaneously identifiable, even for infinitely large phylogenies, they suggest to instead estimate identifiable quantities such as the pulled speciation rate [lambda]_p or pulled diversification rate r_p.
Still later:
Finally, we fit r_p directly with spline functions, as suggested by Louca & Pennell2 (SI S.3), and deduce [lambda] (or [mu]) when assuming [mu] (or [lambda]) constant. General diversification trends are consistent with those found above; finer dynamics depend on the choices made to estimate r_p and suggest interesting new hypotheses to test (SI S.3). Here we chose [lambda] or [mu] to be constant to simplify the problem. Data-driven approaches have relaxed this hypothesis while still estimating global temporal tendencies accurately, based on the reasonable prior belief that rates don’t change much in a small amount of time. We expect such or similar regularization approaches to eliminate the pathological cases with markedly different diversification histories shown in Louca & Pennell.
Lastly, for their paper:
The implications of these results for diversification analyses, however, are largely overinterpreted, mainly because the constraints imposed by the practice of hypothesis-driven research, prior knowledge, and the possibilities to penalize complexity are not considered in the paper. A very similar identifiability issue occurred 10 years ago in population genetics . . . . Identifiability issues naturally arise in approaches that try to infer the potentially unlimited complexity of historical processes from limited contemporary data, and this is why we work hypothesis-driven, develop regularization techniques, and integrate other data types.
Louca and Pennell respond to Morlon, Hortig and Robin: From Nature:
Time-calibrated phylogenies comprising only extant lineages are widely used to estimate historical speciation and extinction rates. Such extinction rate estimates have long been controversial as many phylogenetic studies report zero extinction in many taxa, a findingin conflict with the fossil record. . . . Consequently, estimation methods tend to converge to some scenario congruent to (i.e., statistically indistinguishable from) the true diversification scenario, but not necessarily to the true diversification scenario itself. This congruent scenario may in principle exhibit negative extinction rates, a biologically meaningless but mathematically feasible situation, in which case estimators will tend to hit and stick to the boundary estimate of zero extinction. To test this explanation, we estimated extinction rates using maximum likelihood for a set of simulated trees and for 121 empirical trees, while either allowing or preventing negative extinction rates. We find that the existence of congruence classes and imposed bounds on extinction rates can explain the zero-inflation of previous extinction rate estimates, even for large trees (1000 tips) and in the absence of any detectable model violations. Not only do our results likely resolve a long-standing mystery in phylogenetics, they demonstrate that model congruencies can have severe consequences in practice.
Sounds like evolutionists are left with a lot of 'guessing' to do if they want to publish results. Maybe there's a fairer view of where things stand than this. Here's how ENV summed it up:
Models, which are common in science, are useful but not always realistic. Assumptions and prior beliefs can cloud a hypothesis-driven research project, providing unwarranted confidence in a model that may not have anything to do with true history. It’s also a warning to any scientist, including proponents of intelligent design, to beware of letting assumptions cloud one’s conclusions. Look how modelers can say, “Well, we know a negative extinction rate is meaningless, so we’ll just round it up to zero.”
Where did they go wrong, Orthomyxo? How did they not understand these papers?PaV_{January 21, 2021
January
01
Jan
21
21
2021
11:00 AM
11
11
00
AM
PDT}

Orthomyxo: I wonder if you would agree with Pagel's additional take on Louca and Pennell's original paper:
Amid this epistemological carnage regarding what we can possibly know, the authors helpfully offer some consolation by showing that it is possible to estimate a parameter they call the pulled speciation rate, or lambda_p. This measures the rate of change (the slope of the curve) of the deterministic model of the lineage-through-time plot. The pulled speciation rate can be compared between lineages, or at different times, and might be useful for understanding the processes that gave rise to the species that are alive today, even if not necessarily providing information about those species that didn’t make it. And this aspect — the ones that became extinct — is the deeper lesson of Louca and Pennell’s work. Without fossils, all evolutionary scientists, whether studying speciation and extinction or attempting to reconstruct the features of distant ancestors, need to be aware that the evolutionary processes they identify are those that operated in the species that would survive and eventually leave descendants in the present. We can’t be sure what was going on in those that went extinct. It is the evolutionary version of the observation that history is written by the victors. The supreme irony of this predicament is that Charles Darwin’s idea about the survival of the fittest, the story that we want to understand, by its very nature renders elusive some of the key components needed to study it.
IOW, without the "fossils"--or, to put it into a Darwinian perspective, without a "complete fossil record," 'extant' timetrees are unreliable. That's what it sounds like to me. Keyword: "elusive."PaV_{January 21, 2021
January
01
Jan
21
21
2021
10:24 AM
10
10
24
AM
PDT}

Orthomyxo: I make comments because your comments don't make sense. You say that this whole issue is about "identifiability," and not "trees," but you don't make any clarification about what this might mean. April 2020, commenting on the original paper, Mark Pagel, in the pages of Nature says:
But, writing in Nature, Louca and Pennell1 challenge a major aspect of that enterprise. Specifically, their work regards the issue of estimating past rates of speciation and extinction, which are, respectively, the rates at which new species arise and existing species go extinct. These rates determine the number of contemporary species of various forms. There are, for instance, around 6,600 species of songbird (passerines), which constitute more than half of all existing bird species, and we might therefore be tempted to say that songbirds have a high rate of speciation in comparison with that of other birds. But it’s also possible to speculate that they have a low extinction rate. Louca and Pennell show that the uncertainty is even worse than this: not only can we not estimate these two rates, but also there is an infinite number of different sets of these two parameters that are equally good at describing any particular outcome, such as the number of species of contemporary songbird.
Am I suppose to believe that ENV got it all wrong and I should just take your word for it? I would hope you would explain the issue better. We're open to reason here, but you have to give us something that you've reasoned through--or, at least do it for us.PaV_{January 21, 2021
January
01
Jan
21
21
2021
09:55 AM
9
09
55
AM
PDT}

You are doubling down on the position the following sentences are not about tree-building or comparing trees? "Two theorists have caused a stir in evolutionary circles, claiming to have proven that Darwinian phylogeny efforts (tree-building) cannot be constrained to one “best” answer. In fact, any proposed tree is no better than an infinity of other trees" This after claiming in 23 and elsewhere that the study could be used to compare trees?orthomyxo_{January 21, 2021
January
01
Jan
21
21
2021
09:39 AM
9
09
39
AM
PDT}

What the hell is wrong with you? ENV does NOT say the paper is about estimating time trees. ENV was explaining why the inferences of time trees are as they are in the paper.ET_{January 21, 2021
January
01
Jan
21
21
2021
08:14 AM
8
08
14
AM
PDT}

Again, what the hell is wrong with you? You are back to claiming the sentences "Darwinian phylogeny efforts (tree-building) cannot be constrained to one “best” answer. In fact, any proposed tree is no better than an infinity of other trees." do not refer to tree-building or trees. You can't just deny simply facts out of existence.orthomyxo_{January 20, 2021
January
01
Jan
20
20
2021
04:32 PM
4
04
32
PM
PDT}

Oh my. Yes, they use trees and ENV was pointing out that the trees are nonsense, anyway. I never said the trees were the result. Neither did ENV. They were just pointing out the obviousET_{January 20, 2021
January
01
Jan
20
20
2021
03:20 PM
3
03
20
PM
PDT}

You seem to want to pick nits. Why don’t you simply say that the paper is about “diversification scenarios”? But what are the “diversification scenarios” used for? To build “trees.”
No they are not. They are results unto themselves, allowing you to say things like "speciation in this linage was increased at this time" or perhaps "speciation rates are higher in hot climatic periods". Why do you keep insisting on make comments on a topic you know nothing about?orthomyxo_{January 20, 2021
January
01
Jan
20
20
2021
02:07 PM
2
02
07
PM
PDT}

Orthomyxo: Which paper are you referring to? The original paper--the cause of said controversy, says this in their abstract:
Here we clarify the precise information that can be extracted from extant timetrees under the generalized birth–death model, which underlies most existing methods of estimation. We prove that, for any diversification scenario, there exists an infinite number of alternative diversification scenarios that are equally likely to have generated any given extant timetree. These ‘congruent’ scenarios cannot possibly be distinguished using extant timetrees alone, even in the presence of infinite data.
You seem to want to pick nits. Why don't you simply say that the paper is about "diversification scenarios"? But what are the "diversification scenarios" used for? To build "trees." This solves nothing. The point remains: "for any diversification scenario, there exists an infinite number of alternative diversification scenarios that are equally likely to have generated any given extant time TREE." One divided by infinity is zero. There is Zero probability that ANY scenario is the correct one. Do you want to simply sweep this under the rug by saying it's a matter of "identifiability"? Well, if there are an infinite number of possible diversification scenarios, then all the "identifying" you want to do will get you nowhere. They say: These ‘congruent’ scenarios cannot possibly be distinguished using extant timetrees alone, even in the presence of infinite data. That is, even if you had an infinite amount of data, you cannot "identify" the correct scenario. Will you admit this is a problem?PaV_{January 20, 2021
January
01
Jan
20
20
2021
02:02 PM
2
02
02
PM
PDT}

No daft, just informed. As the quote says, the methods studied in this paper use trees. As in trees are input to the analysis, not the result. You might have saved yourself time and embarrassment of you'd read my first post which explains this.orthomyxo_{January 19, 2021
January
01
Jan
19
19
2021
10:50 AM
10
10
50
AM
PDT}

Ortho:
The paper isn’t about researchers finding flaws in how scientists build trees of life.
Are you daft? Did the researchers find flaws in how scientists build trees?
“Our finding casts serious doubts over literally thousands of studies that use phylogenetic trees of extant data to reconstruct the diversification history of taxa, especially for those taxa where fossils are rare, or that found correlations between environmental factors such as changing global temperatures and species extinction rates,” Louca said, using a term for populations of one or more organisms that form a single unit.
Yes, they did.ET_{January 19, 2021
January
01
Jan
19
19
2021
06:19 AM
6
06
19
AM
PDT}

DNA doesn't determine biological form so any tree made from comparing DNA doesn't have anything to do with reality.ET_{January 19, 2021
January
01
Jan
19
19
2021
06:16 AM
6
06
16
AM
PDT}

That's a long way to say you don't know anything about this topic. The paper is not about ways to estimate trees. The ENV very clearly thinks it is (see the first paragraph). There isn't an infinite number of ways to construct the trees because.... I don't know how many times I need to say this: the paper is not about tree-building.orthomyxo_{January 18, 2021
January
01
Jan
18
18
2021
08:29 PM
8
08
29
PM
PDT}

Orthomyxo: My reply above contained two statements. The first had to do with your charge that the author/s of the ENV thread did not understand what the article at issue was about because the ENV author/s thought is was about "estimating" when, in fact, it wasn't. I pointed out that you're understanding of the ENV author/s view of the article was in error. That was not their take. The second point had to do with your omission of the term "identifiability." My impression is that these kinds of "trees" is sort of your specialty, and that this issue of "identifiability" is your take on the dilemma the paper's article were pointing to. My further point was this: using a word like "identifiability", actually a word describing a mathematical dilemma, doesn't side-step the problem being pointed out: namely, if there is an "infinite" number of ways of constructing the "trees", then the odds of one being right are 1 divided by infinity, which is zero. The problem isn't going to go away, I'm afraid. And it's a 'black-eye' to Darwinian notions of evolution. Now, that's my view and opinion. I'll leave it at that.PaV_{January 18, 2021
January
01
Jan
18
18
2021
03:02 PM
3
03
02
PM
PDT}

Convergent evolution is not relevant to this topic, as it's not about estimating trees. It's also not a major problem for molecular phylogenetics --the reason we know convergences exists is because they stick out as being distinct from the "background" phylogeny shared by the rest of a genome. If we routinely relied on indivudla protein sequences to estimate trees they might be an issue. But we don't . We use DNA, which s much less prone to convergence, and as many genes and sequences as possible to estimate trees.orthomyxo_{January 18, 2021
January
01
Jan
18
18
2021
01:58 PM
1
01
58
PM
PDT}

Orthomyxo, another quote from the University of Oregon website: "An alternative approach that relies on signals of identifiable changes in an organism's genetic makeup also can be misleading." Have you noticed my previous post on convergent evolution (@3) ? Any comments on that ?martin_r_{January 18, 2021
January
01
Jan
18
18
2021
03:27 AM
3
03
27
AM
PDT}

The paper isn’t about researchers finding flaws in how scientists build trees of life. The contents of the paper are rather more important than the title of the press release.orthomyxo_{January 18, 2021
January
01
Jan
18
18
2021
12:25 AM
12
12
25
AM
PDT}

So "Researchers find flaws in how scientists build trees of life" isn't about researchers finding flaws in how scientists build trees of life?ET_{January 17, 2021
January
01
Jan
17
17
2021
08:09 PM
8
08
09
PM
PDT}

Why? Why would actually take the time to argue about something you have no knowledge of? That is not an inference you can draw from the paper, because the paper is not about estimating trees.orthomyxo_{January 17, 2021
January
01
Jan
17
17
2021
05:04 PM
5
05
04
PM
PDT}

Oh my. The inference, from the paper, is that "any proposed tree is no better than an infinity of other trees". Then ENV goes on to explain why THAT is.ET_{January 17, 2021
January
01
Jan
17
17
2021
04:09 PM
4
04
09
PM
PDT}

1 2 Next

You must be logged in to post a comment.

Leave a Reply