I wrote up a critique of an article DI mouthpiece Casey Luskin wrote regarding avian influenza back in October. I don't know whether Luskin ever read my post; at the time, trackbacks to the DI site weren't working. But I'd guess I'm not the only one who pointed out the abundant mistakes in his article, which advanced the thesis that avian influenza wasn't a good example of evolution. He has since written a response to critics here (warning: .pdf file), correcting one of his errors in the original article (and making a confusing mess out of things).
Luskin's original thesis was that H5N1 wasn't a good example of evolution because, he claimed, it was simply a reassortant virus: an avian-human hybrid. Therefore, the "evolution" was not any "new information," but simply a move of information that already existed. Only, of course, the H5N1 strain circulating *isn't* a reassortant virus: it's a pure avian virus. You might think that this tidbit of information would shoot down Luskin's whole thesis, but no, he struggles on.
Though Luskin admits he was wrong when he called H5N1 a reassortant, he still claims:
...regardless of whether or not the Avian Flu has yet evolved into a more deadly form, this evolution would represent small scale genetic change and would not represent an impressive example of evolution.
Contrast this with his original claim:
The reason that the Avian Flu is succeding thus far is because when the two previously-existing viruses swapped some genetic material and created Avian Flu strains, its current configuration is different enough from microbes our immune systems can already target that many people are unable to fight off the virus.
But it's evolution within limits, and it's evolution that generally uses pre-existing genetic material.After all, the current strains of the Avian flu are nothing more than viruses, which are descended from nothing more than a line of billions upon billions of generations of viruses, which, as far as we can tell, have always been viruses, and aren't becoming anything other than more viruses.
Viruses are masters at taking what already exists and swapping it around to dodge our immune system. And that's what has happened here. It's still a virus, and there's probably nothing "new" in terms of new genes. This does not show that evolution can create new genetic information.
Luskin also fails to correct his mistake about the 1918 virus, still claiming it, too, was a reassortant virus (as the viruses that caused pandemics in 1957 and 1968 were). It wasn't. Taubenberger et al. finished sequencing the virus and published their results in October (prior to Luskin's original article on the topic, I might add, and certainly long before his January correction piece). I wrote a bit about the results of the 1918 sequencing study here, and the New England Journal of Medicine has a review here. Quoting from that review:
The startling observation of Taubenberger et al. was that the 1918 virus did not originate through a reassortment event involving a human influenza virus: all eight genes of the H1N1 virus are more closely related to avian influenza viruses than to influenza from any other species, indicating that an avian virus must have infected humans and adapted to them in order to spread from person to person. Thus, pandemic influenza may originate through at least two mechanisms: reassortment between an animal influenza virus and a human influenza virus that yields a new virus, and direct spread and adaptation of a virus from animals to humans.
So Luskin's claims about reassortment in the 1918 virus are still all wet. He's simply swapped one mistake for another.
Luskin continues,
...my critique of the relevance of the evolution of virus by reassortment to supporting Neo-Darwinian claims of macroevolution is still valid.
Now, that's interesting. First, I don't know anyone who would claim that influenza virus evolution was an example of macroevolution, so that's a nice little strawman right there. As Luskin so brilliantly notes in his original article, "in the end, they're still always viruses." Nevertheless, antigenic shift--this swapping of viral segments--is only one way the virus can evolve. What's already been shown to occur with H5N1 is antigenic drift--the accumulation of point mutations in the viral genome. It's these small mutations that require us to re-formulate influenza vaccines every year. Could this drift result in a pandemic strain of virus? There's no reason to discount the possibility, and as I mentioned above, this seems to be the way the 1918 virus evolved. The H5N1 strain circulating already has some point mutations that were detected in the 1918 virus genome. Is it becoming more human-adapted? If so, are these changes an "increase in information," as I asked in my previous post (not addressed in Luskin's correction, of course).
The thing is, even small mutation events can produce slightly modified proteins with novel functions. What the intelligent design folks (and other creationists) keep dodging is whether this novel function represents an "increase in information" or not. If so, what exactly then *does* constitute such an increase? They've denied that gene duplication is such an increase--but such duplication can result in reduced selection pressure on these duplicated genes, eventually resulting in the production novel proteins as mutations accumulate. Is this then an "increase in information?" If not, why not? As they currently define it, there is no way for information to increase. Therefore, even when we give a dramatic example of evolution of a new ability (such as this one), it is denied due to the all-encompassing "increase in information" excuse. It's a "heads I win, tails you lose" situation.
So, once again I will quote from the DI's evolutionnews site:
The misreporting of the evolution issue is one key reason for this new site. The newsmedia in the U.S. seem to have rediscovered the evolution controversy recently. Unfortunately, much of the news coverage has been sloppy, inaccurate, and in several cases, overtly biased.
and wait for Mr. Luskin to do some *real* correcting of his errors regarding influenza virus evolution. I ain't holding my breath.
- Log in to post comments
How many bits of information were there before, Casey? And how many bits are there now?
This is the kind of thing that so mightily pisses me off about the ID movement's flacks: they are willing to lie about stuff that has lethal consequences. Who cares if people die if we don't understand the evolution and epidemiology of H5N1 if God's work is done? It's of a piece with the HIV denial of Johnson and Wells, the suggestion by Deborah Owens-Fink of OBOE that "global warming" should get the same distorted "critical analysis" treatment as evolution in Ohio schools, and the dependence on abstinence-only sex education to keep teenagers out of the back seats of cars. Reality has no place in these people's ideology.
Until he sees a virus turn into a cat-dog, his story will remain the same.
I've just connected a couple of dots. You'll be glad to know that this has nothing to do with ID (or any anagram of ID).
Epidemiologically, the situation with H5N1 is very similar to what's been happening in barley mildew (Blumeria graminis f.sp. hordei, or Erysiphe graminis when I was a lad). In the 1970s a new resistance gene, mlo, was introduced. Normally resistances to mildew are broken down in 3-5 years (for reasons that we understand pretty well). But this gene was different, and has remained effective since then (it's durable, to use the jargon). Up to a few years ago, only two aggessive pathotypes had been found, one was from a breeding experiment, and another which was just wierd. But neither had full virulence: growth went from 4% up to about 20% of normal growth.
Now, as any epidemiologist knows, you only get an epidemic if R_0>1, i.e. if a pathogen population can replace itself. For the mlo aggressive pathotypes, R_0<1, hence no epidemic. This meant only limited selection for mlo aggressiveness (because the population would die out anyway, so it doesn't matter if it takes 2 or 10 generations. There's more to it than this, but I think the essential message holds). And so, aggressiveness doesn't build up, and the resistance isn't broken down.
Finally (phew) we come to the connection to H5N1. If it requires several mutations to become virulent on humans, and for R_0>1, then things might be better than we fear. The mutations will accumulate largely by drift, and this may be slow enough that the molecular biologists will be able to find a vaccine (ha! That will have to be slow :-)).
Someone told me last summer that mlo aggressive isolates were being found in the wild, but the resistance hadn't broken down. And that's after 20 years of use. So, if the large piles of speculation built into the argument above hold, we may be able to solve the problem simply by carrying on as we are destroying habitats until we force all the bird populations to extinction.
Bob
Are viruses as different among one another as, say, mammals differ among mammals? I've heard it said that "a horse is a horse, of course" in reference to the well-document variations through time of the equine lineages (but try to saddle up a Hyracatherium and see how far you ride). So if "a virus is still a virus" then maybe a mammal is still a mammal, and Homo has evolved no further than Australopithecus.
It's the old creationist 2-step...if you show evidence of small scale changes they dismiss it as "only microevolution", and if you show the evidence for the large scale changes they demand to see the step-by-step micro-scale processes.
As for "information", that word is usually left vaguely defined by the ID community. Problem is, every time you actually use a technical definition of "information" complete with metric (and they do exist), evolution has a maddening way of increasing same. Hence they have to equivocate with the definition.
Mark,
It can vary widely, but within a species, there's generally much more divergence within a microbial species than within a mammal species. For example, 2 isolates of E. coli can share as little as ~60% of their DNA. Look at what a difference only ~2% difference between humans and chimps makes. So, yeah, it's silly to say, "but it's still a virus." Tiny genetic differences can make huge differences in clinical presentation (as you'll see from a post I'm working on...)
Are viruses as different among one another as, say, mammals differ among mammals? I've heard it said that "a horse is a horse, of course" in reference to the well-document variations through time of the equine lineages (but try to saddle up a Hyracatherium and see how far you ride). So if "a virus is still a virus" then maybe a mammal is still a mammal, and Homo has evolved no further than Australopithecus.
Good point mark.
It would seem no matter how different 2 viruses (or two bacteria) we must consider them the same (micro-scale variants) according to the creationists based on the fact we call them both 'viruses'.
But if we consider a flu virus verses a bacteriophage, clearly there are major differences and one is not a simple variant of the other. If something like a flu virus evolved into something like a bacteriophage, wouldn't that be considered a macro-evolutionary event?
You'd think the DI would hire someone slightly less clueless about evolution than Luskin as their mouthpiece, maybe, heaven forfend, a biologist!
Even if Luskin were correct (he is not), to take his article seriously one would first have to agree that the word "god" is the same as the word "dog," and that the sentence "Frank killed Joe" is equivalent to "Joe killed Frank."
What a fool.
I can't help but notice that, in Luskin's response to his critics, the only source he cites for scientific information is Wikipedia.
Tara, perhaps you could prepare a brief essay on what would happen if the folks over at the DI got control of science in the US, especially with respect to emerging diseases. That might go over well as a letter to the editor, or-- better still-- as a piece in a magazine with a nice broad readership.
People need to understand that what they don't know *can* hurt them.
Maybe we should all go rewrite the Wikipedia articles so that they contain massive amounts of evolutionary data, and thus "contaminate" the DI's source of "scientific information"?
One thing I've always wondered is why these people, with all their money, do so much blogging, editorializing, and pop-culture bookselling, rather than technical article writing? They can't get it into a journal? Why the hell can't they make their own, online journal for the articles?
Oh wait, I forgot, this is really about the Renewal of Science (and Culture), not advancing science itself, nevermind...
Let me confess I'm your culprit: Whenever these ID folk claim that evolution of influenza isn't "real" evolution because it's still the same species, I tell them otherwise. Am I wrong to do so? Is my shorthand error for evolution as bad as their shorthand errors that completely screw things up?
Heck, correct me if I'm wrong -- I'll change my tune, but sadly -- but influenza viruses are species specific. Pig viruses give flu to pigs, human viruses give flu to humans, avian viruses give flu to birds (and sometimes just some birds, leaving others unaffected). Southeast Asia has long been a hotbed of new, nasty viruses that feature material from birds, pigs and humans, because in that area (say, along the Mekong River) humans live over the water, close to pigs and waterfowl. In these environs viruses can sometimes swap material.
Generally the viruses stick to the species they torment. Leaps do occur however. One of the issues in a leap from birds to humans is that in birds the infection is passed through fecal droppings -- the virus affects the gut more than other organs. In humans, the virus passes through exhalations from the lungs -- the virus affects the lungs and UR tract more than other organs. Since virus "species" generally don't encompass both ends, such a shift would indeed suggest a new species.
Am I in error? What is the terminology I want?
BTW, some years ago (1980?) human vaccinations nearly wiped out camels across southern Asia. Turns out the human vaccine strain easily jumped to camels, and it was very deadly to them. Flu vaccines are not tested to be sure they won't jump victim species.
Help me out on this -- I don't want to be the one who makes a Casey Luskin argument accurate. I don't have to be so careful in the law, I don't think; I don't think he'll ever get it right. But I'm years out of the lab, years out of biology. What is the proper thing to call a brand new flu virus? Is it just a "daughter?"
1. It would difficult to find among the Eukaryota differences as great as those between the bacteriophage f1 and HIV. But they're both "viruses". What's in a word?
2. As to spontaneous increase of information: Consider a megamole (~ 10^31) of random, transient molecular strings or constructs of a small number of distinct units that can assemble and disassemble over moderate timescales. Posit that there is no preference among the possible combinations, so the entropy of the population is high. If by chance one string happens to acquire the property of self-replication, then rapidly the pool of component elements will be directed into that distinctive non-random form, and it will come to dominate the population. The entropy of that local system will have decreased; information will have been spontaneously "created." Is there any reason in principle why that can't happen? If not, then acquisition of a new function (or phenotype) upon random mutation of a vial genome that allows its more successful replication in a particular environment is, by analogy, an increase in information. In the case of the flu virus, adaptation to a new environment of humans (vs. birds) is the non-random selective agent leading to gain in information.
If the IDers have a different idea about how information is gained, such as implantation by some supernatural intelligence, let them say so explicitely.
1. It would difficult to find among all Eukaryota differences as great as those between the bacteriophage f1 and HIV. But they're both "viruses". What's in a word?
2. As to spontaneous increase of information: Consider a megamole (~ 10^31) of random, transient molecular strings or constructs of a small number of distinct units that can assemble and disassemble over moderate timescales. Posit that there is no preference among the possible combinations, so the entropy of the population is high. If by chance one string happens to acquire the property of self-replication, then rapidly the pool of component elements will be directed into that distinctive non-random form and it will come to dominate the population. The entropy of that local system will have decreased; information will have been spontaneously "created." (No violation of 2nd Law!) Is there any reason in principle why that can't happen? If not, then acquisition of a new function (or phenotype) upon random mutation of a viral genome that allows its more successful replication in a particular environment is, by analogy, an increase in information. In the case of the flu virus, adaptation to a new environment of humans (cf. birds) is the non-random selective agent leading to gain in information.
If the IDers have a different idea about how information is gained, such as implantation by some supernatural intelligence, let them say so explicitely.
(sorry for double post)
it seems like Luskin is often relying on a catch-all facile argument: any number you care to nominate isn't new information because it can be constructed from all those other numbers randomly lying around. Apparently there is no such thing as information.
Well now, Luskin does have one good case in point for his information arguments:
as we have seen, DI statements over many years have still failed to include any increase in information. Maybe this is some sort of existentialist method of illustrating their point?
Tara writes: "As they currently define it, there is no way for information to increase."
This shouldn't be surprising. It's their entire point. Remember, things are "designed". Therefore, only the "Designer" can add new information by creating a new "kind". It is impossible, by definition, for new information to "spontaneously" appear. Therefore, anything that merely looks like new information must, by definition, *not* be new information. It's up to right-thinking apologists to explain how it is not new information. It's kind of like Flood Geology. Yeah, it only *looks* like millions of years of silt deposits.
b.spitzer writes:
"One thing I've always wondered is why these people, with all their money, do so much blogging, editorializing, and pop-culture bookselling, rather than technical article writing? They can't get it into a journal? Why the hell can't they make their own, online journal for the articles?"
They have an online journal. Here's the latest issue:
http://www.iscid.org/pcid/2005/4/2/pcid_contents_2005_4_2.php
Don't get your hopes up. For instance, Dembski has an article in this issue, "Information as a Measure of Variation". You can read virtually the same thing at:
http://www.designinference.com/documents/2004.08.Variational_Informatio…
dating back to July, 2004.
Of course, this issue of PCID is pretty tame compared to past issues, which features some out and out no possibility of denying it cranks. Check out the second issue of PCID at:
http://www.iscid.org/pcid/2002/1/2-3/pcid_contents_2002_1_2-3.php
and feast your eyes on Chris Langan's "The Cognitive-Theoretic Model of the Universe: A New Kind of Reality Theory ". Put your Crank-O-Meter in safe mode first or you'll bend the needle.
I honestly don't know why Casey doesn't contribute to the journal. He'd fit right in.
Ed--
Sure, that's one way, but it's not the only way they can evolve. Additionally, research done at the Center for Emerging Infectious Diseases here suggests that people in close contact with pigs may be infected more often than we realize.
I'm not 100% sure what you mean here. H5N1 has actually been reported to cause problems at "both ends," so to speak (see this reference)--is that a new "species?"
Not exactly sure. :) This is one thing that makes discussing "species" when referring to viruses and bacteria rather difficult. Obviously we don't have parameters that are used in sexually reproducing populations, such as reproductive isolation, so we generally look for common features and genetic sequences.
Interesting..I'd not heard about the camels before. And you're correct, they're not tested on other species, and it would be very difficult (impossible?) to do so. Imagine all the permutations...
We generally just refer to it as a new variant, and cite the serotype. In the literature, they're named by their site and year of isolation. For example, one influenza A virus found in New Jersey in 1976 is called A/New Jersey/76, and is a H1N1 virus. This gets a bit bogged down into details, though--"variant" is easier.
Hmmmm... I seem to recall that it was some kind of disease, but I hadn't heard that it was caused by human vaccinations. How did they ensure that the new camels they imported from Australia didn't get sick the same way? Did they stop using the vaccinations?
Thanks for the explanation, Tara. I'm not convinced I've not given Luskin a BB to shoot back, but I know a lot more.
Camel flu: Try this article I found at PubMed: