I come from the Michael Behe school if ID. I accept common descent, by which I mean universal common ancestry. It seems to be the consensus view in science, it seems reasonable to me, and I don’t have any compelling reasons to doubt it.
But could I actually defend my belief in common ancestry if asked? All living organisms share certain features in common. Organisms leave offspring. Therefore, universal common ancestry. That sounds pretty weak, I admit. I need to do better.
I don’t believe that new organisms appear out of thin air, so to speak. I accept that the organisms of today are the offspring of prior organisms. Even young earth creationists accept common descent prior to the flood and common decent after the flood, though they resist the idea of universal common ancestry. I still haven’t figured out how and where they draw the lines though, so why should I draw a line that I cannot defend. Therefore common ancestry. That still sounds pretty weak, I admit. I still need to do better.
So time to hit the books. What are the arguments for universal common ancestry in the books on evolution that I have. Erm… Halp!
Why do folks believe in universal common ancestry and what is the best book on “the evidence for evolution” to refer to in order to find the best arguments in favor of universal common ancestry?
What if it’s not the evidence for the theory of common descent per se that leads people to accept it, but rather how the theory of common descent provides an explanatory framework. Does its explanatory power outweigh the need for evidence for it and do people confuse what it can explain with what constitutes evidence for the theory?
What is the evidence for universal common ancestry and why is it considered evidence for universal common ancestry?
Stop digging, Alan.
Erik, let’s talk about some common descent that isn’t universal, just a few species in one group. Let me suggest an example that I’m intimately familiar with:
Harshman J., Braun E.L., Braun M.J., Huddleston C.J., Bowie R.C.K., Chojnowski J.L., Hackett S.J., Han K.-L., Kimball R.T., Marks B.D., Miglia K.J., Moore W.S., Reddy S., Sheldon F.H., Steadman D.W., Steppan S.J., Witt C.C., Yuri T. Phylogenomic evidence for multiple losses of flight in ratite birds. Proceedings of the National Academy of Sciences 2008; 105:13462-13467.
Do you think it presents evidence that those species are related by common descent?
Alan Fox,
Strictly, in PCR at least, one adds the polymerases as proteins from some other source than the molecule being copied, so I can see it being objected to on that technicality. It is a good example of the template-driven nature of the process though.
John Harshman,
Thank you (sincerely) for the suggestion, but first I will chew through Theobald’s response to K&W and I will also read K&W directly. It’s a lotta fun so I will take my time with it. Theobald’s first article was no fun at all.
Erik,
Yeast cells replicate their DNA when they reproduce, just like bacteria. Again, this is news to you?
Erik,
Sheesh. I am sure you are not thick, but you certainly make it hard to sustain that view at times. I didn’t expect you to look them up before I mentioned them. I mentioned them, then suggested that you may care to look them up … you want spoon feeding, that I see. But your response to any effort anyone makes in that regard is to act like a dick.
Yeast? All cells self-replicate in the way I meant – by template-directed replication of their DNA, whose sequence includes the sequence of the proteins responsible for that template-directed replication. There is a layer of complexity provided by sex, although yeast are sexual anyway. And by multicellularity.
Whoah. And you think this does not undercut your inference (that speciation and common descent are a real thing)? Anyway, for the time being I must leave in the air the fact that when I asked for an observation, you responded with an inference, not being quite straightforward at first that it’s no more than an inference.
What caught my attention was your phrase “self-replication of DNA”. Conceptually, self-replication is a radically different thing than mere replication. Now I see you have changed the story: cells self-replicate while DNA only replicates. This requires no special attention.
As you were.
Allan Miller,
And ubiqutous! And observed!
Erik,
You actually thought that explaining sequence similarity required not just replication of DNA, but self-replication, unaided by any cellular machinery???
Dang, Erik.
Yes, I do. Why would it? First, that inference does not come mostly from fossils but from the nested hierarchy found among extant species. Second, as I have pointed out several times so far, we don’t need to be able to identify a fossil as an ancestor or even identify species in order to infer common descent. We can use individual specimens if we have to, and we can analyze them without any assumptions of ancestry or species membership.
You toss around “inference” as if it’s a bad thing rather than the foundation of all science. Observations are raw data. Raw data are useless if you can’t make inferences from them. I gave you both the data and the inference from them. “No more than an inference” is what all knowledge is; the question to ask is how strong the inference is. You clearly have a fundamental misunderstanding of the nature of science.
How do we measure the strength of an inference, and how strong does an inference need to be in order to become a conclusion?
I mention that because you are denying any common descent of species whatsoever, not just universal common descent. We really should get down to those lower levels. We may have to go to lower-level phylogeny just to discover your limits.
Anyone know of a good multi-gene primate phylogeny paper?
We’re watching you. 🙂
In other words, it comes from “we can build a tree”. And if the analogy with linguistics applies in any relevant sense, that’s exactly backwards. This is why I hope to see in the Theobald debacle some good indication how biologist build trees (“phylogenetic inferences”) and why they think it proves what they think it proves.
No, it does’t come from “we can build a tree”, which distorts and trivializes what I said. (That’s not a very honest method of argument, by the way.) A better statement would be “all different sorts of data give the same, consistent signal, for which there is no other explanation than a tree”. Universal common descent is not really the best introduction to tree-building, especially if what you’re really wondering about is whether any species at all are related. Though the methods are more or less the same.
Which is exactly how it works in linguistics – multiple lines of relevant data (relevant in the sense that organic and borrowed material is carefully distinguished) must point to the same source to justify drawing the tree. Yet linguists don’t think the tree in itself proves that languages evolve. What proves that languages evolve is the history of observations long before anyone began drawing language family trees. Biologists have the last point backwards.
Your conclusion doesn’t follow from your premises. Just because there was earlier evidence that languages evolve, that doesn’t mean that later work is not also evidence. It could in fact be better evidence. Historical priority doesn’t determine quality of evidence.
Now of course there is evidence of common descent that precedes tree-drawing. Faunal progression in the fossil record is an example of that. It’s just that the very strongest evidence, all that genomic stuff, is quite recent. Nor does that mean that the evidence we had earlier wasn’t sufficient for a strong conclusion. It was. Still, if you want the very best, go with the molecules.
Here’s another paper that might be simpler to understand, because the sequences are short, the species even more closely related, and some of the methods are more intuitive:
Harshman J., Huddleston C.J., Bollback J., Parsons T.M., Braun M.J. True and false gharials: A nuclear gene phylogeny of Crocodylia. Systematic Biology 2003; 52:386-402.”>Harshman J., Huddleston C.J., Bollback J., Parsons T.M., Braun M.J. True and false gharials: A nuclear gene phylogeny of Crocodylia. Systematic Biology 2003; 52:386-402.
True, later language trees are also evidence – because we know from prior empirical evidence that that’s the sort of data that indicates a language family, so we can draw a tree.
Linguists draw a tree because the languages are related. Biologists say a group of animals are related because they can draw a tree.
So the evidence has to be based on evidence. How do you know that the prior empirical evidence is sound, if you don’t have evidence for that evidence?
See, there are two things that occur in science that you don’t bother with–generalization, and cause-effect analysis. We can generalize empirically across language and manuscript evolution to other types of evolution to recognize that nested hierarchies are characteristic of and diagnostic for descent with modification. Related, but not actually the same thing, we can figure out that if life is derived from prior life–even if it has changed somewhat or even a lot–that we’d have to end up with the derivative patterns that we see in nested hierarchies.
It’s utterly absurd to claim that we have to see something happen in order to infer certain causes from their expected effects. Using your reasoning, we can trust that footprints in the dust are produced by humans and other animals, while footprints on an asteroid couldn’t be understood as being due to life, since we’ve never seen the aliens (or whatever) that made them.
The fact is that certain effects point to certain causes, and, unless you have a sound reason to think that something quite different did it, one is entitled to believe that those certain effects were due to those certain causes.
Glen Davidson
Again you seek to trivialize my point by distorting it. The fact you can draw a consistent tree of languages shows that they are related. If they were not related, you couldn’t draw a consistent tree. The only explanation for the ability to draw a consistent tree is relationship. You may have known they were related already, but in fact that knowledge was superfluous to the task. If somehow you had picked unrelated languages to try making a tree from, you would have failed to produce a consistent tree from the data.
Now, what’s wrong with the sort of data phylogeneticists use that makes us unable to infer common descent from our ability to draw consistent trees?
Look, you are the one trivializing and distorting my point here, not the other way round. And this is a problem that evidently will not go away. I guess I must live with it.
Right, but the fact the you can draw a consistent tree, does NOT in itself show that the languages are related. There is a standard of evidence that justifies the tree. Sometimes the evidence is so scant and distorted that even though you can draw a consistent tree, the tree is not justified. Note: Language isolates are a thing.
So we can infer certain causes just by way of generalization? What sort of certain causes are we permitted to infer without observation and why speciation and common descent belong to that certain category?
Do you even try to understand what is written?
Glen Davidson
Yes it does. In fact the two statements “if they were not related, you could not draw a consistent tree” and “if you can draw a consistent tree, they are related” are logically equivalent to each other. NotA implies NotB = B implies A.
Of course language isolates are a thing. And you can’t draw a consistent tree linking language isolates. That’s why they’re isolates. Perhaps the confusion lies in the meaning of “consistent”, by which I mean you consistently get the same tree from non-overlapping data sets.
We can infer causes that are the only explanation of the data, or at least that explain the data very much better than alternative causes. Again, this is universal in science. How do you know that carbon atoms have 6 protons? By observation? Of course not. Nobody has ever seen a proton, much less the protons in a carbon nucleus. How do you know that the protein sequence of insulin in chimps and humans is the same? By observation? Of course not. Nobody has ever looked at insulin and seen the amino acids. And so on. You clearly have no idea how science works.
First, you are not in position to teach me linguistics.
This out of the way, one last time how it really works. Relatedness of languages is a matter of evaluation of the data that is, as a matter of empirical historical knowledge about languages (from which the linguistic theory has been extracted), known to demonstrate relatedness. The tree would be an illustration of that relatedness, not a demonstration or proof of it. There are other illustrations of that data too, such as maps of isoglosses. The most common (and best) illustration is tables of grammatical features.
Important note: Since we know what sort of features demonstrate relatedness and what sort of features don’t, a table can also illustrate unrelatedness! A language tree is the worst kind of illustration in linguistics because it barely allows the representation of the relevant data on it. The tree is a high-level abstraction only after the data has underwent every other sort of analysis. The same way as a few tables are not by themselves a demonstration of relatedness (as said, they could serve the purpose of illustrating unrelatedness), neither is the tree.
The demonstration or proof is fundamentally the empirical historical knowledge about what languages are and how they change. People have been making observations about languages throughout history and they have been seen to branch into daughter languages, dialects, to pidginize, etc. – and that’s the proof, not any sort of tree. Biologists lack such empirical knowledge about speciation and common descent. They have only the tree drawn on anatomical and genetic data. The fact that biologists must tie the cart before the horse like this due to necessity has no effect on other sciences where the subject matter is different.
You’re spending most of your time trying to get people out of jail I hope. You don’t have a moral excuse for spending even a second on this forum. Good bye Erik.
There’s no difficulty understanding your insults. Or John’s.
Erik seems not to get logic, specifically the “if…then” form of thought. Perhaps he thinks that it only works with deduction, and that with the lack of certainty it won’t do. Well, it’s not of deductive certainty in science, of course, but it does work so long as one understands that the conclusions are tentative, and that there’s always a chance that something else did it.
It’s just a fact, if life arose via common descent, then the “tree of life” and nested hierarchies. Were one able to logically predict that if life arose by design, by leprechaun magic, or any other unknown “cause,” then the “tree of life” and nested hierarchies, there would be another good candidate (at least until the progression of life and transitionals come in), only there isn’t another good candidate for the patterns of life.
If the fingerprint on the gun is Mac’s, and the gun is what caused the murder (by ballistics evidence), then Mac is the murderer. That’s how we think, appropriately, and according to most contexts (yes, Mac could be innocent because of certain things (he would have touched the gun later, for instance), but we’re talking the typical circumstances). If the nested hierarchies, common descent. Anything else requires special pleading, or Erik’s constant attempts to avoid typical scientific inference.
It’s not perfect, we know, but it is hugely better than any other way of getting to the (apparent) truth of life, earth, and the universe.
Glen Davidson
You mean John’s “NotA implies NotB = B implies A”? Fill in what he meant by A and B and we’ll see about the logic. Remember that he was theorizing about linguistics at that moment.
In linguistics, the prior evidence is the historical record – empirical observations of linguistic phenomena, change, branching, convergence, etc. This is as sound as it gets.
Agreed. I am, however, apparently in a position to teach you logic, if you meant what you said, or perhaps how to compose an English sentence, if you didn’t mean what you said. I also seem to be in a position to tell you how science works, since you don’t seem to know.
Tables really aren’t very good illustrations of relationships among languages, certainly not of branching relationships, which is what we’re talking about here. Nor are a lot of language families provided with historical knowledge. I think we may have surpassed the limits of your ability to explain things in English. But I can derive at least one correct statement from this that relates to phylogenetics. The tree itself is not a demonstration of relatedness. It’s the fact that the data force a particular tree, and that different data sets force the same tree, that’s the demonstration of relatedness.
Not true. At most it can illustrate that there is no evidence of relatedness. It’s certainly possible that Basque and Polish are related. It seems in fact quite likely to me, as human language certainly precedes the exit from Africa. But there is no evidence of that relatedness.
I’m sorry, but I find that nonsensical. The demonstration is in the data, i.e. the cognates of various sorts. General knowledge of “what languages are”, whatever that means, and how they change might help guide the analysis, but it doesn’t demonstrate anything.
Nonsense. This sort of observation is not available for most languages and most times. It’s nice to have but it isn’t necessary. Let’s remember Algonquian.
Of course we have such empirical knowledge. The raw data of molecular phylogenetics are sequence differences, and we know how those differences arise — through mutation — and we can observe (a trickier word than you seem to imagine) those differences arising within populations in the present. Speciation isn’t actually something we need to know about in phylogenetics, but we can observe various steps and intermediate states of it in the present too. And we can see fossils that show stages of transformation from one sort of species to another. All of these are reasonably analogous to the sort of thing you have in linguistics.
Now what you want to see, apparently, is a movie of a cow changing into a whale, but that’s your problem, not biology’s. Nobody can show you a movie of Latin changing into French either, but you don’t care about that.
Erik,
If one were to adopt your own approach to this evidence, one would require an actual observation of one language changing into another. A few words here or there are simply not good enough – by your own standards.
Do you realise there is an equivalent historical record in genomes? It’s imperfect, being subject to the vagaries of time, but it is there.
I made a statement, you agreed that it was true, and then you claimed that a logically equivalent statement was false. The problem there was your ability to reason, not the subject.
Eric may disagree about what “historical record” means. But it occurs to me to wonder whether the evidence of experimental phylogenies would give him the sort of evidence he demands. If you google that term you will get access to the literature on them, the first that I know of being this one:
http://science.sciencemag.org/content/255/5044/589
Briefly, one works with fast-evolving sequences of clones, checks the sequences of populations frequently, and removes individuals to new media frequently to produce a branching tree. Then one tests phylogenetic methods against that known, real-time phylogeny. Hey, they work!
If you want to teach me logic, then why not do it by filling in what you meant by A and B above? And if you agree that you are not in position to teach me linguistics, then why do you keep doing it?
In contrast, I am in position to teach you linguistics just fine and here it comes again.
You have evidently never seen how it works. It’s done by lists of cognates (words with their meanings) with their historical reconstruction by internal and comparative analysis. That’s tables.
So according to you there are two kinds of languages: Demonstrably related and undemonstrably related. How very Darwinian of you. How about the theory (fact rather) that, even though e.g. French and HIndi share a common history, they don’t seem to have anything much in common now. Wouldn’t this be even more true of languages that have no demonstrable common history? Even though reality is more complicated than that, for example areal convergence without historical organic relation is very common. So there’s grounds to posit more kinds of languages – organically related, areally related and not related in either of the aforementioned ways.
You mean those observations don’t exist? Feel like reading a book?
That which you think doesn’t mean anything and what you find nonsensical is the historical record of the fact that languages change. When I say “what languages are” I mean the observed reality that languages are changeable and there is a long record of observations of how the changes occur. Those observations serve to justify that languages with scant historical record can be treated in a similar manner – that they are changeable and there are certain rules to how they change. Because in other cases it’s been observed that that’s how languages behave and that’s what they are.
Empirical reality seems to be too much over your head. That’s way too sad.
So there’s no evidence for evolution. But that’s not biology’s problem. Got it.
John, to Erik:
Erik,
You’re making a fool of yourself in this discussion, largely because you do not grasp the point John is making in the quote above.
It’s remarkable to me that you could read through Theobald’s 29+ Evidences without getting it, but you have surprised me with your incomprehension before.
In any case, here’s a suggestion for you. Why not go off, read Theobald again, and think things over until you actually get what John is saying there, and can explain it in your own words?
You don’t have to agree with it, but be able to present it in a way that demonstrates your full understanding, free of your habitual strawmanizing tendencies.
You’ve heard of the “Ideological Turing Test”?
There must be some reason I took keiths off Ignore. I think I’m going to blame DNA_Jock for that. Fortunately, it can be rectified. 🙂
No, failure of logic or reading comprehension there Erik. He’s not saying they’re “undemonstrably related”, as in thought to be related with no evidence.
He’s saying there are languages for which we can’t say, given the evidence available, whether they are related or not. But they still could be, we just don’t know that.
Moved a comment to guano
I should not need to remind long-time members gratuitously insulting fellow members is against site rules.
Better don’t go there, unless you look forward to really long discussions about the definition of “a language”.
You know, like: “Yes it changed, but it is still the same language!!!” or philosophical mesmerizing about instantations of ideal languages.
No. Harshman says the following, “It’s certainly possible that Basque and Polish are related. It seems in fact quite likely to me…” In other words, undemonstrably related!
Perhaps the concept of a language changing over time but being described, by convention, as the same language, would help with temporal boundaries regarding evolving species.
But not demonstrably unrelated!
Alan Fox,
Whatever happened to “NotA implies NotB = B implies A”? If the criteria that are known to determine relatedness between languages fail to determine relatedness between some given languages, it should be legitimate to conclude that those given languages are demonstrably unrelated.
Erik,
Haha! Mr Can’t-be-arsed-with-genetics wants me to read a book! The evidence available in books of ‘observation’ of language evolution mirror pretty closely those of ‘observation’ of genetic evolution. You dismiss one, you dismiss ’em both. Accept one, accept ’em both. Some of your co-religionists would have it that languages were all ‘created’ anyway, Tower of Babel stylee.
Mung,
So absent a movie there’s no evidence? Mad as a box of frogs, the lot of youse!
Allan Miller,
You mean both are inference, like JH says?
Thanks for the heads up
Erik,
I don’t know why you have a problem with inference. ID is one big inference. The Tale Of Sal And The Spliceosome is inference. Your take on global Flood stories is inference. Of course, not all inferences are created equal.
If there were a movie, it would take too long to watch. Ergo, they’d demand some other sort of evidence, which would also fail for reasons.
Glen Davidson