When I’m bored, I’ll periodically take a look at the blogs published by
the bozos at the Discovery Institute. I can generally find something good for a laugh. So I was doing that tonight, and came across yet another example of how they try to distort
reality and use slimily dishonest math to try to criticize the evidence for evolution. This time, it’s an article by “Logan Gage” called What exactly does genetic similarity demonstrate?.
Francix X. Clines, an excellent writer for The City Life and Editorial Observer sections of The New York Times, today (April 23, 2007) repeats what may be the most common mistake in trying to sell Darwinism to the public. In “Evolution, on Broadway and Off,” Clines writes of the American Museum of Natural History’s exhibition on evolution:
The DNA exhibit shows how the chimpanzee’s DNA has been conclusively shown to be 98.8 percent the same as the visitor’s DNA. Hey, that’s no show stopper for the monkey-song chorus — it still allows a one in 100 chance they’re right.
In other words, you are silly for not believing in Darwinism because you have very similar genes which make the proteins in your body as the chimps do to make their proteins. Game over, right? Not so fast.
You can see right there, from just the first paragraph of Gage’s writing what kind of scam he going to try to pull. What’s the evidence for common descent? Is it just that one figure – the 98.8% similarity between chimpanzee and human genes?
Of course not. He’s trying to take the number out of context, because out of context, it’s a whole lot easier to come up with arguments against it. In context, it’s one data point among many: as we sequence the genes of different creatures, we find
that the similarity relations between different species for a tree, where the more similar two species’ genes are, the closer they are to one another in the tree. And that tree is an almost perfect match for the tree that was developed without genetic comparisons based on other kinds of evidence.
So – a perfectly nested tree of genetic relationships, which matches predictions made before anyone had any idea of how to sequence the genome of a species. And what’s Gage’s response to that?
Second, the 98.8% DNA sequence similarity between chimps and humans that Clines references does not even establish claim one (common ancestry). And “you don’t have to take my word for it,” as LeVar Burton always used to say on Reading Rainbow.
As Francis Collins, head of the project which mapped the human genome, has written of DNA sequence similarities, “This evidence alone does not, of course, prove a common ancestor” because an intelligent cause can reuse successful design principles. We know this because we are intelligent agents ourselves, and we do this all the time. We take instructions we have written for one thing and use them for another. The similarity is not the result of a blind mechanism but rather the result of our intelligent activity.
Some design proponents think the evidence for common ancestry is good (e.g., Michael Behe), while others–citing the fossil record, especially The Cambrian Explosion–do not. But neither group thinks that sequence similarity alone proves either common ancestry or the Darwinian mechanism, as so many science writers of our day seem eager to assume.
See? The out-of-context number is used as if it’s the only evidence that exists. How do you think that Collin’s quote really ends? No surprise: with a lead-in to how much other evidence there is in addition to that one single number.
So it’s just more dishonest quote mining from the DI. Big surprise, eh?
As is pointed out by Ken Miller in a number of lectures which can be downloaded, even more convincing is the identification of two ape chromosomes which merged to form one chromosome in the genome of homo sapien sapiens. This finding came out of the sequencing of the human and chimpanzee DNA. One doesn’t hear the ID clowns commenting on this remarkable coincidence.
I think that we have 40% kinship with a loaf of bread ( wheat genome ) 🙂
This is a standard creationist lie. They can’t explain nested hierarchy, so they simply lie and portray it as simple “similarity,” often as “the argument from similarity,” which I, as a molecular biologist, have never heard of.
I suggest that whenever we are treated to this lie, we ask, “I don’t know of anything called ‘the argument from similarity,’ but I do know about nested hierarchies. Nested hierarchies are a lot more than simple similarity. How do you explain a nested hierarchy, when no sets of designed objects fit into nested hierarchies that are superimposable on the nested hierarchies derived from examining their components?” or “Is it honest to describe a well-defined mathematical relationship as nothing more than ‘similarity’?”
The ultimate irony is that the only figure in The Origin of Species is a nested hierarchy; they are obsessed with calling us “Darwinists,” but they are avoiding a prediction that Darwin made.
Just a general comment
If you can make maths fun and interesting for me – i’ll sell you one of my livers.
Cheers,
Ché
It seems to me that the Intelligent Design liars could equally well argue that, even if the Chimpanzee genome and the Human genome were 100% identical, the two species neither have to be the same, nor have common ancestry.
If “The Designer” — the Wizard of Oz, Thor, or Squamous Sirians in flying saucers — wanted the chimpanzee different from the human, why let mere DNA stand in the way?
They could insist that Translation and Transcription are merely theories…
I agree that this use of selective statistics and implicit dismissal of evidence by using the term “similarity” is intellectually dishonest. Incomplete descriptions are not satisfactory.
I also think that to address these arguments it is best to consider an anti-evolutionist viewpoint. Correlation (similarity) does not imply causation, and they make this point. From their viewpoint a creator would certainly reuse portions of the genome, for rational reasons; why reinvent the wheel. Any intelligent creator would also use these building blocks in a nested hierarchy. I naturally use nested hierarchies for programming.
The essay does use this statistic outside the context of neo-darwinism, just as a darwinist speaks of genomic evidence outside the context of creationism. A creationist might complain that a scientific argument doesn’t consider other “evidence” for intelligent design. Calling it a lie does not address the creationist context, some people really belive in extra-physical causation.
Any intelligent creator would also use these building blocks in a nested hierarchy.
You’re gonna need to back that up, maybe find some way to refute my argument that ‘any’ intelligent designer would use identical (not just similar) genes to perform identical functions.
Any intelligent creator would also use these building blocks in a nested hierarchy.
I’m not sure nested hierarchies mean the same thing in biology and computer science just because they use the same word. But I’ll let you have that out with the biologists and the programmers.
edit, second paragraph should be quoting
I naturally use nested hierarchies for programming.
Nested hierarchies in computer programming have nothing whatsoever to do with the genetic/biological term nested heirarchy.
A better analogy from computer science would be imagine the diff history of a source code control system. You can imagine forking a project at some point, and continue with two different projects. Diffs that occur before the fork are shared, after the fork, they’re different unless both happened to evolve in parallel (fixed the same bug, perhaps). The two branches are a nested heirarchy in a biological sense; it has nothing to do with the structure of the source language.
One interesting difference between a computer program in this sense and a biological system is what happens to deeply important structures. An intelligent designer can completely refactor some part of a computer program, but that never happens in genetics – highly conserved sequences turn out to be important, and important sequences turn out to be highly conserved.
Showing evidence of significant refactoring without an evolutionary path would be evidence for ID.
Is it just me, or is anyone else imagining that Ché has got a cellar stacked with jars of assorted body parts?
Mark Dow:
“Any intelligent creator would also use these building blocks in a nested hierarchy.”
Utterly, totally wrong. No intelligent creator has ever done so. Remember that the components have to fit a single nested hierarchy that is superimposable on the nested hierarchy of the assemblies. Can you wrap your brain around this essential concept?
Here are some more: we can separate functionally-relevant from functionally-irrelevant differences between sequences and get the same hierarchy. We can separate domains of homologous proteins and get the same hierarchy, unless we happen to cross a domain-shuffling mutation.
“I naturally use nested hierarchies for programming.”
Yes, but when you do so, the components (commands) don’t fit in a superimposable nested hierarchy, and they don’t have functionally irrelevant differences that fit the same hierarchy. They also don’t have partially-overlapping functions as protein families do.
“The essay does use this statistic outside the context of neo-darwinism,…”
There weren’t any statistics in the essay.
“… just as a darwinist speaks of genomic evidence outside the context of creationism. A creationist might complain that a scientific argument doesn’t consider other “evidence” for intelligent design.”
Science isn’t fundamentally about “considering evidence,” it’s about making and testing predictions; the nested hierarchies are an absolutely required by common descent. Scientific controversies are resolved by data, not essays and critiques.
“Calling it a lie does not address the creationist context, some people really belive in extra-physical causation.”
It is a lie in the creationist because Collins didn’t write that for the reasons that the dishonest Gage claimed he did. Collins wrote it because there is far more evidence than mere similarity to consider. Gage doesn’t consider all the evidence, and neither do you.
There’s a reason why our judicial system has immense disdain for hearsay, and it is the same reason why creationists and IDers use it in lieu of citing evidence–they intend to deceive their audiences.
Drekab:
“I’m not sure nested hierarchies mean the same thing in biology and computer science just because they use the same word. But I’ll let you have that out with the biologists and the programmers.”
The term “nested” means exactly the same thing. But my understanding that it is more commonly used to modify “loop” in programming jargon.
IDists really think that God is stupid! (Or at least not perfect.)
I just had a thought. The ID folks insist that evolution alone isn’t good enough to explain our current biological diversity, so God has to “tamper” with it to get a desired result. So does that mean that God was too stupid to set up a creative mechanism that would right the first time?
typo on the last line
should be “…would work right..”
Zombi:
Point taken, and I agree that a serial programming analogy does not map well to genetic evolution.
Just to be a devil’s advocate, in the same sense of my original post, with respect to your comment:
“Showing evidence of significant refactoring without an evolutionary path would be evidence for ID.”
Yes it would, and I would desire some test for this. ID advocates present “evidence” of this all the time (“how can an eyeball be incrementally designed” and “what good is half a wing”). The scientist has the problem that they can’t show conclusive evidence that significant refactoring did not occur, particularly without complete source code history. But presumeably God is a very clever programmer, and his refactoring does not leave the same tracks as my hacks.
Collins is no evolutionary biologist, and has been heavily criticized for his book. One can stop there and engage the argument instead.
However, I note the humor in that creationist quote mining has become so bad that they can be used to refute each other. In a review by another creationist:
( http://www.curledup.com/langogod.htm )
So what Collins really says is that observations of DNA sequencing only follows from a common ancestor, ie as predicted by a theory of “common descent with modification”. Now, he makes this assertion grudgingly, but he does make it, refuting Gage’s mining.
Zombi:
Point taken, and I agree that a serial programming analogy does not map well to genetic evolution.
Just to play devil’s advocate, in the same sense of my original post, with respect to your comment:
“Showing evidence of significant refactoring without an evolutionary path would be evidence for ID.”
Yes it would, and I would desire some test for this. ID advocates present “evidence” of this all the time (“how can an eyeball be incrementally designed” and “what good is half a wing”). The scientist has the problem that they can’t show conclusive evidence that significant refactoring did not occur, particularly without complete source code history. And presumeably God is a very clever programmer, and his refactoring does not leave the same tracks as my hacks.
Torbjorn:
“Collins is no evolutionary biologist, and has been heavily criticized for his book.”
So what? He hasn’t been criticized for that part of it. This is a ludicrous attack on Collins, as we geneticists have generated gigabytes of this sequence data. We understand nested hierarchies just as well, or better than, evolutionary biologists, as we have been constructing them for decades.
“One can stop there and engage the argument instead.”
There’s no argument there. Just a gross misrepresentation.
“…ID advocates present “evidence” of this all the time (“how can an eyeball be incrementally designed” and “what good is half a wing”).”
That’s not “evidence.”
“The scientist has the problem that they can’t show conclusive evidence that significant refactoring did not occur, particularly without complete source code history.”
We don’t have that problem, because we have conclusive evidence in the form of the nested hierarchy. We don’t need “complete source code history,” as we have a huge sample of the existing source code, and statistics tells us that our p values for our hierarchies are ludicrously small. The NH also makes clear predictions about what we will find as we continue to look at existing “source code,” and those are fulfilled hundreds of times a day.
“And presumeably God is a very clever programmer, and his refactoring does not leave the same tracks as my hacks.”
The tracks are all there in the nested hierarchy, Mark. Perhaps you should look at some data before pontificating.
Um, where is the correlation? Evolutionary theory predicts nested hierarchies caused by diverse mechanisms. So by confirming the prediction we prove that the causality happened.
Perhaps you mean that since we don’t observe the nested hierarchy evolve, our confirmation involves correlation between an observation and a prediction? Sort of like the usual argument about historical sciences, I suppose.
It feels like a category mistake, or at least a proof problem. And indeed, I’m pretty sure most modern measurements of causality has this problem, we can’t track every constituent in real time. My radio is causal, not correlated, thank you. 🙂
Usually, to get from mere correlations to a firm suggestion of causality, we change a parameter and observe if the process changed as well. For example, by changing radio channel, I discover that my seemingly correlated radio follows other stations as well.
We can (and do) that here as well on the phenomena “now”, so we have fulfilled that criteria. And in fact observed the causality itself.
But more directly here, the nested hierarchy itself fits a very few trees with any appreciable likelihood in the vast space of possibilities. In fitting such a precise pattern to such a huge precision, or conversely constrain the subset of possibilities that correlate to a very few, I think the practical difference between a correlation and a causality is pretty non-existent in this specific case.
(In the larger view, we can dispel much of the “historical” mythos altogether, since we observe the phenomena “now” as well, we can still make repeatable observations, et cetera.)
Well, since creationists traditionally refuse to describe their creator and his workings, they can’t really make that prediction.
But they can certainly claim that their theory can be correlated with such an observation. 😉
Pickled liver? Ergh!
But we can humor Ché with noting that if he sells one of his livers he will still have a set of them – the empty set.
“And presumeably God is a very clever programmer, and his refactoring does not leave the same tracks as my hacks.”
Refactoring doesn’t mean the before-and-after are bad. They just have to be different, and as we know, what works in one organism in one environment isn’t so great in another organism in a different environment. We have tools for comparing how different (minimum edit distance would seem to apply to genetics just like it does lines of text). These tracks would be recognizable. And while we may not have a complete history, we have many versions derived from different common sources.
Take the classic retinal detachment thing in mammals. In a very early version, it doesn’t matter which way the optic nerve attaches to the retina. In later versions, it matters a lot. Why didn’t God change it?
The only explanation ID has for this is that in every single possible case where God could have refactored or rewritten the source code, and even ought to have done so, he for one reason or another, chose not to. We can find no example, macroscopic or microscopic, which has the recognizable fingerprint of intelligent intervention characterized by a significant, singular change from one implementation to a different one.
John:
“The tracks are all there in the nested hierarchy, Mark. Perhaps you should look at some data before pontificating.”
I was hoping not to be associated with the pontificate, by identifyng with the devil:)
The tracks are very convincing to me (for evolution by natural selection), and I have no doubt. I was only pointing out that God’s refactoring might leave similar tracks, for reasons we can’t devine. I leave coding tracks that you wouldn’t be able to fully retrace. But, I think this is an empty argument for ID.
John:
If you are trying to take me to task for not having read Collins book, you are correct. That is why I found the above creationist quote instead.
My assumption was that since he made such a mess of where morality is observed and how it can come about, and similar problems with the evolutionary account as noted elsewhere, we could tell the creationists that they shouldn’t criticize evolutionary biology from such a basis to get around which part is or isn’t correct.
There are a lot of textbooks on evolutionary biology that we and them could discuss the theory around instead. Assuming, of course, that they are willing to engage with the theory. 😉
See, that was engaging the argument. Of course, one can be more specific and supportive, as Mark was.
…
The rest of the comment seems to be directed to other commenters. Really, I think you make some good points (since you made the same point on the precision of hierarchies as I did :-), but your comments are a bit scattered. This makes it hard to follow.
I think your point with regards me was that I am attacking Collins unfairly. Well, I have certainly found reason to criticize the small parts I have been presented with. Here however my interest is mostly in commenting on the DI creationism, not Collins’. I would probably have to read his book to say anything new.
But since creationism is anti-science, it is fair game. 😉
Mark:
“The tracks are very convincing to me (for evolution by natural selection),…”
But many of the tracks (“silent” changes) represent evolution by drift, not natural selection. They are just as important in evolutionary theory, and completely unexplainable by ID.
“… and I have no doubt. I was only pointing out that God’s refactoring might leave similar tracks, for reasons we can’t devine.”
But the IDers claim to be able to divine the reasons, and they are lying, because they can’t explain these tracks. We know that hundreds of human homologs will rescue mouse mutants, so why should they be 5% different at the aa level?
“I leave coding tracks that you wouldn’t be able to fully retrace.”
Of course, but someone with your design expertise would see the intelligent reasons for doing so in most cases. That’s not the case in biology. BTW, when I engage in design by directed mutagenesis, I create or destroy nearby restriction sites so that I can distinguish my mutant from endogenous wild-type easily–if I didn’t do that, my colleagues would rightly call me stupid. That’s still nothing like the nested hierarchies that we see in biology.
“But, I think this is an empty argument for ID.”
MarkCC’s point is that not only is it empty, it is dishonest.
The “argument from leaps” can agree evolution, but it could disprove it as well.
It cannot disprove ID, but it could still agree with it.
In other words: The fact we find no leaps (refactors, whatever) does not disprove ID (because ID is consistent with anything), but it [b]also[/b] fails to disprove evolution …
… while if we DID find leaps, it would disprove evolution (or at least the current formulation thereof) while still not saying anything about ID.
This is the issue with, in fact, all tests: ID can never fail, while evolution could but hasn’t.
Therefore, it is more likely that evolution is “right”, because if ID were the case, we’d expect ‘by chance’ that since many of these sorts of tests would have results independent of one another, at least one would turn out in such a way that evolution were disproven – but they haven’t.
From comment #1: As is pointed out by Ken Miller in a number of lectures which can be downloaded…
Could someone post a link to these? Ta.