Comment #31780

Posted by Noturus on May 23, 2005 11:46 PM (e) (s)

I enjoyed this article.  Except for this line:

“A random mutation in an organism, like a random change in any finely tuned machine, is almost always bad.”

Random mutatations are usually neutral, yes?

Comment #31781

Posted by PvM on May 23, 2005 11:56 PM (e) (s)

Correct. Most mutations are neutral. In fact, neutrality seems to be a major contributor to evolvability and actually may be under selective pressure. Imagine that, neutrality can be selected for.

Comment #31782

Posted by speaker4thedeads on May 24, 2005 12:29 AM (e) (s)

The newest dance craze…”The Behe Backstep”.

Comment #31783

Posted by speaker4thedeads on May 24, 2005 12:33 AM (e) (s)

The newest dance craze…”The Behe Backstep”.

Comment #31784

Posted by speaker4thedead on May 24, 2005 12:38 AM (e) (s)

oops. sorry for the double post…swore to myself I wouldn’t do that.

Comment #31793

Posted by Fernmonkey on May 24, 2005 05:22 AM (e) (s)

Also, when Dembski is wielding his equations, he gets to play the part of the hard scientist busily correcting the errors of those soft-headed biologists. (Evolutionary biology actually features an extraordinarily sophisticated body of mathematical theory, a fact not widely known because neither of evolution’s great popularizers—Richard Dawkins and the late Stephen Jay Gould—did much math.)

Oh dear.  Traumatic memories of my undergrad population genetics lectures are now surfacing.  Incredibly smart professor, but a lousy communicator.

Comment #31800

Posted by TonyB on May 24, 2005 08:23 AM (e) (s)

Biologists aren’t alarmed by intelligent design’s arrival in Dover and elsewhere because they have all sworn allegiance to atheistic materialism;

Look out, H. Allen Orr. The dishonest quote-miners in the creationist community (and they are legion) will soon be using the above quote without bothering to cite the words immediately following:

they’re alarmed because intelligent design is junk science.

It’s how the anti-Darwinists work.

Comment #31803

Posted by Nat Whilk on May 24, 2005 09:02 AM (e) (s)

Evolutionary biology actually features an extraordinarily sophisticated body of mathematical theory

Recommendations as to the best first book for a mathematician to read about the extraordinarily sophisticated mathematics of evolutionary biology?

TIA

Comment #31807

Posted by Russell on May 24, 2005 09:19 AM (e) (s)

Tony B. underestimates the creationist quote miners. I predict that

H. Allen Orr wrote:

]Biologists aren’t alarmed by intelligent design’s arrival in Dover and elsewhere because they have all sworn allegiance to atheistic materialism;

will be transcribed as:

H. Allen Orr wrote:

Biologists… have all sworn allegiance to atheistic materialism;

Comment #31810

Posted by Steve Reuland on May 24, 2005 09:49 AM (e) (s)

It’s a good article, but here are a few things that bugged me…

1.  Orr unfortunately falls for the old trick that the IDists accept an old Earth and some form of evolution.  This is not the case.  As I’ve pointed out here many times, the ID movement takes no position on these issues.  This crucial distinction is necessary for understanding the unscientific nature of the ID movement.

2.  The biggest problem with Dembski’s “specified compexity” is that under no circumstance has he presented evidence, matematical or otherwise, that it cannot evolve naturally.  To make matters worse, Dembski defines SC ambigiously.  He uses not only the definition that Orr does, but also defines it as something which has less than a 1 in 10^150 chance of occuring by any natural means.  By using both definitions, Dembski engages in circular reasoning.  Orr does a good job of pointing out some of the problems with SC, but at its core, the argument is entirely question begging and requires no special refutation.

3.  Orr gives the IDists a complete pass on their relationship to the old-school creationists.  Even Behe and Dembski’s arguments are not really new, but can be found in various forms among earlier creationists.

Comment #31813

Posted by Steven Thomas Smith on May 24, 2005 10:00 AM (e) (s)

Recommendations as to the best first book for a mathematician to read about the extraordinarily sophisticated mathematics of evolutionary biology?

• Ewens, Mathematical Population Genetics. 2d ed. Springer-Verlag, 2004.

• Felsenstein, Inferring Phylogenies. Sinauer, 2003.

Comment #31814

Posted by Flint on May 24, 2005 10:03 AM (e) (s)

Orr might have also mentioned that CSI has never been subjected to a test, and that repeated demands of the form “here’s an item. Does it have CSI?” have been ignored. It should be safe to say that CSI cannot be calculated for any item unless the answer is known in advance.

Comment #31816

Posted by Andrea Bottaro on May 24, 2005 10:25 AM (e) (s)

Dembski’s arguments have been met with tremendous enthusiasm in the I.D. movement. In part, that’s because an innumerate public is easily impressed by a bit of mathematics. Also, when Dembski is wielding his equations, he gets to play the part of the hard scientist busily correcting the errors of those soft-headed biologists.

Ironically, Dembski’s original argument (by his own admission) is a wholesale application of  Fisherian statistics, which was developed in the context of evolutionary population genetics three quarters of a century ago.  As evolutionary mathematics goes, Dembski’s application of Fisher for eliminative induction is rather pedestrian and utterly old-fashioned (and as soon as he leaves Fisher, as in his foray into the NFL theorems, largely wrong).  Any good mathematical evolutionary biology grad student would probably give Dembski a run for his money on these topics.

Comment #31822

Posted by Ed Darrell on May 24, 2005 11:42 AM (e) (s)

I finally realized why I’ve been so uncomfortable with Dembski’s use of the “no free lunch” stuff:  Intelligent design assumes there IS a free lunch, provided by some other intelligent entity.

Or do I completely misunderstand the the claim of NFL?

And, Andrea — can we get some of those “good mathematical evolutionary biology grad” students to follow Dembski around like a good Truth Squad, to give Dembski a run for his money?  The more the merrier.

Comment #31831

Posted by Steviepinhead on May 24, 2005 12:23 PM (e) (s)

I’m going to make every effort to be at Dembski’s Seattle lecture tonight, but I’m neither a working biologist nor a good mathematical evo-bio grad student, merely a humble defense attorney.  No anxiety throwing some of the Lenny Flank-type questions his way, if I get any reasonable chance, just for fun, but if any of you have more specific suggestions for great one-liners, I’d love to see them.

Remember, given the usual tactics and the likely partisan nature of the audience, I don’t anticipate getting more than one good shot, at most, so please feed me that one killer question or comment you’ve always wanted to toss old Bill’s way…!

Comment #31834

Posted by Stuart Presnell on May 24, 2005 12:53 PM (e) (s)

Steve Reuland wrote:

The biggest problem with Dembski’s “specified compexity” is that under no circumstance has he presented evidence, matematical or otherwise, that it cannot evolve naturally.

Wading into that argument is, I fear, just falling for the trap Dembski & Behe deliberately set by introducing all this CSI fog.  We say “You can’t prove that SC systems can’t evolve naturally”.  They cry back “Aha, but you can’t prove that it can evolve naturally”.  The ignorant insufficiently educated public watches this tennis match, with little guidance as to whose court the ball resides in.  Who should get the benefit of the doubt?  With whom does the burden of proof rest?  That’s the kind of confusion Dembski and co. want to engender, because it moves the debate away from science, where they’re weak, and onto rhetoric, where they might stand a better chance.

The great battle for the Benefit of the Doubt!  Scientists point to science’s track record of success, and demand the benefit of the doubt on any as-yet unanswered questions.  “Give us long enough”, they say, “and we’ll figure out how [insert IDist’s favourite unexplained biological phenomenon] evolved.”  ID advocates (those who aren’t just recycling the old canards, at least) say “They’ve had long enough!  So many years of trying, and still they can’t answer these questions!”  And of course they always have the vast resource of Newton’s “great ocean of truth” to draw upon, an unbounded set of unresolved mysteries to throw at us.

The anti-evolutionists’ backup plan has to be to undermine the definition of science itself within the curriculum.  We saw hints of this in the recent Kansas business.  Really, it’s been implicit all along in Dembski’s writing.  The only way for his “Explanatory Filter” to be anything other than an Argument from Ignorance is to change the rules of science itself: to flip the benefit of the doubt away from science, and to throw out methodological naturalism to admit supernatural explanations.  As their direct assaults on evolution lose traction with the public (as the coverage of Kansas suggests they will), we can expect more of this Plan-B philosophical fancy footwork instead.

Comment #31836

Posted by Zippypinhead on May 24, 2005 12:57 PM (e) (s)

I’m going to make every effort to be at Dembski’s Seattle lecture tonight, … given the usual tactics and the likely partisan nature of the audience, I don’t anticipate getting more than one good shot …

Take your pick:

Why not ask why Dembski doesn’t retract his proven lies about the scientific literature?

Orr went real easy on Dembski in the New Yorker — he could have quoted  the mathematician David Wolpert:

David Wolpert (No Free Lunch Theorems expert) wrote:

“William Dembski’s treatment of the No Free Lunch theorems is written in jello”

Dembski on the scientific failure/PR success of ID to date:

Dembski wrote:

“We have done amazingly well in creating a cultural movement, but we must not exaggerate ID’s successes on the scientific front.”

Dembski on Christian motivations of ID:

Dembski wrote:

“intelligent design is just the Logos theology of John’s Gospel restated in the idiom of information theory.”

The American Association for the Advancement of Science’s statement on ID says:

“the ID movement has failed to offer credible scientific evidence to support their claim that ID undermines the current scientifically accepted theory of evolution”

More on Dembski here.

Comment #31839

Posted by Longhorm on May 24, 2005 01:07 PM (e) (s)

According to Orr,

According to Darwinism, evolution largely reflects the combined action of random mutation and natural selection.

I don’t know how Orr is using the term “Darwinism.”  But Charles Darwin didn’t know about genes, mutations, genetic recombination or what happens at the genetic level when sexual reproduction occurs.  Maybe Orr was referring to the way a lot of people use the term “Darwinism.”  But a couple important points.  Probably most mutations were not harmful.  Most mutations did not make populations of organisms less apt to keep on reproducing.  That is not to say that, even over the long term, most mutations tended to help the population of organisms keep on reproducing.  Most mutations were probably trivial in terms of reproductive fitness.  Kimura had the idea that most mutations were neutral or only mildly deleterious.  How would one characterize the mutation that causes achondroplasia (dwarfism)?  My understanding is that, under some conditions, the mutation can be harmful if both parents have that gene.  But little people still reproduce. 

One interesting thing to remember: amphibians evolved from fish, but there are still lots and lots of fish on earth.  A population of organisms might have a mutation that helps members of the population reproduce over millions of years.  But organisms without that mutation may also reproduce over millions of years. 

Also, according to the geneticist John Drake, RNA-based lytic viruses average 1 new mutation per division, and bacteriophage M13 averages .0046 new mutations per division.  Humans average 1.6 new mutations per sexual generation among coding genes; mice average about 1 new mutations per sexual generation among coding genes; C. elegans (worms) average .036; fruit flies average .14.  I tend to think that if most mutations had really been “harmful,” a lot of populations of organisms wouldn’t have been around as long as they have been.  Also, humans may average 50 to 150 mutations per sexual generation among non-coding genes.  My understanding is that DNA that didn’t code at one point did sometimes code at a subsequent point. 

Orr suggests that organisms coming into being with new mutations was the main proximate cause of phenotypic difference from one generation to the next.  Obviously that is true for asexually reproducing organisms such as bacteria.  But it is not true for sexual reproducers.  Among sexual reproducers, the main cause of difference from one organism to the next is sexual reproduction.  Sexual reproduction always has resulted in the offspring having a different genotype and phenotype than either of its parents had.  Not massively different — squirrels don’t give birth to armadillos.  But always a little different.  I’m quite different than my parents.  I suspect that most of the differences among dogs were caused by sexual reproduction and not mutation.

There is a misconception that a lot of people have (even Darwin had it) that sexual reproduction is a blending process.  It’s not. There is no blending like putting two cans of paint in a trough.  What happens is that a series of cell-divisions results in sex-cells.  They have half the chromosomes of normal cells.  And the chunks of DNA are in different orders than are the chunks of DNA in regular cells.  When the sperm-cell fertilizes the egg-cell, the two units of chromosomes don’t blend.  They don’t even touch each other.  They just sit to each other in the nucleus of the cell.  Over massive lengths of time, this process contributed to significant differences among some organisms.  For instance, chihuahuas and saint bernards. 

Many scientists overemphasize the causal importance of mutation and underemphasize the importance of sexual reproduction.  Maybe that is because “mutation” sounds more exotic.  By “mutation” I mean any cell-division other than meiosis in which the daughter-cell has a genome that is different than the genome of the parent-cell.  That is not to say that organisms having mutations was not important in bringing about the existence of, and differences among, many organisms.  It was important.  For one thing, it is the main thing that varies the number of nucleotides in the genome.  My understanding is humans have a larger number (though not significantly larger) number of genes than mice do.  Also, a human being born with a mutation resulted in the first human having blue eyes. 

The causes of mutation are an interesting issue, but basically beyond the scope of this discussion.  But one important: Many mutations have been caused in part by some organisms producing the number of offspring that they did. 

But the combination of genetic recombination, which results in sex cells being the way they are, and sexual reproduction was hugely important in causing the differences between rodent-like mammals and gorillas.  Remember: we are talking massive lengths of time.  And sex results in the offspring being different than their parents.  Look at any litter of puppies.  Even if there is not one new mutation.  That is one reason we care so much about who we reproduce with.  We contribute half of the chromosomes, and our partner contributes half.  The chromosomes don’t blend.  They just sit next to each other.  So who we choose matters.  It has a huge affect on what our offspring is like.

On a different note: People who refer to themselves as proponents of “intelligent design” rarely indicate what event(s) on earth they believe the designer(s) caused.  For instance, did the design turn dust — poof! — directly into two elephants (one male and one female)?  I suspect that most, if not all, of the events they believe the designer proximately caused the designer did not proximately cause.  For instance, I suspect that most of them think a designer turned dust directly into the first two humans (one male and one female).  And that didn’t occur.  The first organisms to live on earth that were fairly similar to me were born by their mothers in much the same way I was born by mine.  Self-replicating molecules evolved (through reproduction) into all the organisms to have lived on earth. 

Also, something’s being complex does not, by itself, enable us to determine that a being turned inert matter (or “nothingness”) — poof! — directly into it.  I’m relatively complex, and I was born by my mother.

Comment #31843

Posted by Longhorm on May 24, 2005 01:45 PM (e) (s)

Steve Reuland posts:

The biggest problem with Dembski’s “specified compexity” is that under no circumstance has he presented evidence, matematical or otherwise, that it cannot evolve naturally.

I don’t think one has an obligation to “present evidence, matematical or otherwise, that [a thing] cannot evolve naturally.”  I’m not sure what that means.  And the claim that “X was designed” is a positive claim, though a vague claim.  But it would be nice if the person who offers that claim says which event(s) he or she thinks the designer caused.  And then present any data that she thinks enables her to determine that.  Many of the claims I’ve seen offered by people who call themselves proponents of “intelligent design” are vague.  Were humans “designed?”  Well, if that means that God turned dust directly into the first two humans (one male and one female), no.  They were born.  Either humans and apes either share common ancestors or they don’t.  And they do.  At least it is overwhelmingly probable that they do. 

Also, does my being complex enable one to determine that a being turned dust directly into me?  No.  In fact, I was born by my mother.

Does the complexity of the universe help us determine that a being caused the series of events that resulted in the matter, space and time that we associate with the Big Bang?  That’s more complicated.  I don’t want to wade into that right now. 

I don’t think I would mind if teachers said that some people believe that a “a being caused the Big Bang.”  I don’t think I would even mind if it were done in biology class as long as they teach evolution and teach it well, and make clear that it occurred, or at least that it is overwhelmingly well-supported.  It would also depend on teacher and the context.  However, cosmology should not be part of the biology curricula.

But I concede that I don’t know the series of events that resulted in the matter, space and time that we associate with the known universe.  In fact, I don’t know the series of events that resulted in self-replicators being on earth.  But self-replicating molecules evolved (through reproduction) into all the organisms that have lived on earth.  Or at least it is overwhelmingly probable that they did.  And we should teach evolution and teach it well.

Comment #31847

Posted by Longhorm on May 24, 2005 02:01 PM (e) (s)

Diet also played an enormous role in bringing about the existence of, and differences among, many organisms.  Diet causes some organisms to live longer.  It makes some organisms bigger.  Bone length depends in part on diet.  That is one reason Latin Americans tend to be shorter.  I’m sure dinosaurs got as big as they did partly because they ate so damn much.  The length of the neck of the giraffe was affected by diet.

If you eat more, you tend to consist of more cells.  Cells divide more frequently if they are well-nourished.  That makes you bigger.

But don’t eat too much or you are going to be out of shape.  Unless you exercise.

Comment #31848

Posted by Arne Langsetmo on May 24, 2005 02:07 PM (e) (s)

Remember, given the usual tactics and the likely partisan nature of the audience, I don’t anticipate getting more than one good shot, at most, so please feed me that one killer question or comment you’ve always wanted to toss old Bill’s way…!

How about:  “Hey Bill, why do you insist that painting the bulls-eye around the arrow after it’s hit a tree is a fair method of scoring?”

If he plays dumb, explain that his calculations (i.e., “scoring” of probabilities) is done after the “desired” outcome has already been specified, and that the a priori probability of even the totally random occurence of the heads-tails sequence “HTTHHTTHTHHTTTHTTHTHHH……” (with a million coin flips) is 1 in 2^1,000,000, but that after you have seen a particular sequence and defined it as the target outcome, the probability that it did occur is precisely 1.

Cheers,

Comment #31849

Posted by Sir_Toejam on May 24, 2005 02:17 PM (e) (s)

“And the claim that “X was designed” is a positive claim, though a vague claim”

You can look at it that way as a linguistic statement and be correct, but it is realistically preposterous to do so. 

If I claim black is white, am i making a positive claim?  sure, liguistically i am, but realistically…

However, if I claim black is white, and give some evidence to back that up (maybe i show that individual variation in visual processing has a lot to do with whether one views a color as black or white), then i would say i am making a positive claim.

besides that…

The primary argument of ID is saying black is not black, it could be white.  That is not a positive declaration, even linguistically.

Comment #31859

Posted by Jim Wynne on May 24, 2005 02:58 PM (e) (s)

Lonhorm wrote:

Diet also played an enormous role in bringing about the existence of, and differences among, many organisms.

At first glance, I read “organisms” as “orgasms” and was going to beg for details, but never mind.

Comment #31865

Posted by Russell on May 24, 2005 03:24 PM (e) (s)

Steve Reuland wrote:

The biggest problem with Dembski’s “specified compexity” is that under no circumstance has he presented evidence, matematical or otherwise, that it cannot evolve naturally.

To which Longhorm responded:

I don’t think one has an obligation to “present evidence, matematical or otherwise, that [a thing] cannot evolve naturally.”

To which I respond:

No, “one” doesn’t have an obligation, but Dembski does. Isn’t that what his “4th law” and entire crusade is founded on?

Comment #31866

Posted by Alan Jenkins on May 24, 2005 03:25 PM (e) (s)

Concerning the argument that most (by the vast majority)mutations are harmful used by the ID/creationist (they ARE the same), I wonder how they would explain the number of alleles represented in the human population for, lets say, the HLA genes?  The HLA-B has, a last count, 108 alleles in the human population. If you take a young earth that creationist propose of less than 10,000 years, and two original parents (Adam and Eve)you will get VERY frequent mutations to create 108 HLA-B alleles in 10K years even if both Adam and Eve were heterozygous for HLA-B and neither shared the same allele for that gene.  Note:I consider a new allele a mutation.  That is 104 new alleles in 400 generations (using 25yr generation time for humans) or roughly 1 new allele every 4 generations.  That is a hell of a lot of mutating!  And the only harmful thing about it is trying to get a damn transplant when you need one!

Comment #31867

Posted by Longhorm on May 24, 2005 03:30 PM (e) (s)

Sir T posts:

You can look at it that way as a linguistic statement and be correct, but it is realistically preposterous to do so. 

If I claim black is white, am i making a positive claim?  sure, liguistically i am, but realistically…

However, if I claim black is white, and give some evidence to back that up (maybe i show that individual variation in visual processing has a lot to do with whether one views a color as black or white), then i would say i am making a positive claim.

besides that…

The primary argument of ID is saying black is not black, it could be white.  That is not a positive declaration, even linguistically.

Let’s say one were to offer the claim: “Humans were designed.”  In most context without further elaboration, I would not be justified in believing that the claim is true.  The claim is too vague.  Now I don’t know that the claim is false.  It’s too vague for me to know that it is false.  But let’s say that, after being pressed, the person who offers the claim were to get more specific.  Let’s say he says: “A deity turned dust — poof — directly into the first two humans (one male and one female).”  Well, it is clear that that did not occur.  Or at least it is overwhelmingly probable that it did not occur.  Let’s just say it didn’t.

But an advantage of vagueness is that you often can’t be sure that the claim is false. 

On a different note, mutations did bring about much of the difference between sponge-like creatures and humans.  Much.  But sexual reproduction also brought about a lot of the difference between sponge-like creatures and humans.  And I feel very confident that sexual reproduction brought about a lot the difference between rodent-like mammals and humans.  And, obviously, some organisms having produced the number of offspring that they did was hugely important in rodent-like mammals evolving into humans.  There are no more australopithicenes around, and I think about 6 billion humans. 

There is also the issue of what kinds of events have caused mutations.  I’m not going to get into that now.  But it is an important issue.  I think more work should be done on it. 

Jim posted:

At first glance, I read “organisms” as “orgasms” and was going to beg for details, but never mind.

Yeah, that would definitely have required further elaboration.

Comment #31869

Posted by Longhorm on May 24, 2005 03:43 PM (e) (s)

Russell posted:

No, “one” doesn’t have an obligation, but Dembski does. Isn’t that what his “4th law” and entire crusade is founded on?

I disagree, though I can’t get into the issue further.  But I do think he is obligated is to say what events the designer caused.  At least some of them.  That would help us determine whether his claims are reasonable. 

For instance, did the designer turned inert matter (or “nothingness”) — poof! — directly into the first two humans?

Comment #31870

Posted by SWSchaeffer on May 24, 2005 03:54 PM (e) (s)

The question about the neutrality of mutations confuses the introduction of mutations with the fate of mutations: 

“I enjoyed this article.  Except for this line:

“A random mutation in an organism, like a random change in any finely tuned machine, is almost always bad.”

Random mutatations are usually neutral, yes?”

One must be careful about defining where the mutations occur in the genome when asking about their effects.  Two thirds of random mutations introduced into the coding region of a gene are likely to be deleterious because of changes to the amino acid sequence of the encoded protein.  These mutations although introduced, are rarely fixed in populations either because of stochastic loss or due to purifying selection against the deleterious change.  Many of the nucleotide changes in noncoding sequences are likely to be neutral, except for those that occur in cis-acting regulatory sequences.  With respect to coding sequences, Orr is correct.

Comment #31873

Posted by Longhorm on May 24, 2005 04:22 PM (e) (s)

SWSchaeffer posts:

Two thirds of random mutations introduced into the coding region of a gene are likely to be deleterious because of changes to the amino acid sequence of the encoded protein.

SW, thanks for the post.  How are you using the word “deleterious?”  And what evidence is that there most mutations are “deleterious?”  And do you have a reference?  Is there an especially good article on the issue?  I hear this discussion a lot, and it seems like there is not a clear consensus. 

In his book The Ancestor’s Tale, Richard Dawkins deals with this issue.  He definitely seems to think that most mutations to non-coding DNA are “neutral.”  As for coding DNA, his position is harder to tease out.  He seems on the fence to some extent.  But I would say that finally he is more sympathetic with your view.  He speaks favorably of Tomoko Ohta’s position, namely that most mutations to coding DNA are “nearly neutral instead of completely neutral.”

Comment #31876

Posted by Arne Langsetmo on May 24, 2005 04:32 PM (e) (s)

Concerning the argument that most (by the vast majority)mutations are harmful used by the ID/creationist (they ARE the same), I wonder how they would explain the number of alleles represented in the human population for, lets say, the HLA genes?  The HLA-B has, a last count, 108 alleles in the human population. If you take a young earth that creationist propose of less than 10,000 years, and two original parents (Adam and Eve)you will get VERY frequent mutations to create 108 HLA-B alleles in 10K years even if both Adam and Eve were heterozygous for HLA-B and neither shared the same allele for that gene.  Note:I consider a new allele a mutation.  That is 104 new alleles in 400 generations (using 25yr generation time for humans) or roughly 1 new allele every 4 generations.  That is a hell of a lot of mutating!

Your answer in formal mathematical notation:

G₀ dI/dⁱt

Hope that explains everything.

Cheers,

Comment #31878

Posted by Steve Reuland on May 24, 2005 04:44 PM (e) (s)

Longhorn wrote:

I don’t think one has an obligation to “present evidence, matematical or otherwise, that [a thing] cannot evolve naturally.”  I’m not sure what that means. 

Dembski claims that specified complexity is evidence for ID.  He claims that when one sees SC, then one should infer design, because SC objects cannot come about naturally.  (If the meaning of “natural” isn’t quite clear, that’s Dembski’s problem — he is the one drawing the dichotomy between “natural” and “intelligent” causes, and claiming to find evidence of the latter by way of rejection of the former.)  Therefore, Dembski is obligated to demonstrate that SC objects cannot evolve naturally.  Otherwise, his argument fails. 

And it’s not anyone else’s obligation to demonstrate that any given thing can evolve naturally, much less provide a specific explanation for how it did evolve.  It is acceptable to say “we don’t know” for the time being.  Not knowing how something evolved is not the same as saying that it couldn’t.

And the claim that “X was designed” is a positive claim, though a vague claim.  But it would be nice if the person who offers that claim says which event(s) he or she thinks the designer caused.  And then present any data that she thinks enables her to determine that. 

And the data Dembski presents is specified complexity.  (Which of course is not data, but an argument that certain sets of data must be the result of ID.)  For a long time, he claimed that “the math” showed that specified complexity could not evolve.  Skeptics asked to see these calculations that suposedly demonstrated that SC objects could not come about by natural means.  Dembski vacillated and kept promising that they were forthcoming (after claiming that they had already been done).  Finally, in No Free Lunch, he presented the one (1) calculation that he has done to date.  This was on the flagellum of  E. coli, for which he made the grand discovery that a random combination of amino acids is extremely unlikely to result in a functioning flagellum.  Since no one thinks that happened anyway, the calculation was pointless.  No evolutionary scenarios were taken into account.  His excuse for this was that irreducible complexity made the evolution of the flagellum impossible, which if true, would be enough by itself to conclude that it couldn’t have evolved naturally.  Thus, Dembski’s SC doesn’t actually add anything of merit.

Comment #31884

Posted by Steve Reuland on May 24, 2005 05:07 PM (e) (s)

SWSchaeffer wrote:

Two thirds of random mutations introduced into the coding region of a gene are likely to be deleterious because of changes to the amino acid sequence of the encoded protein. 

But not all amino acid changes are deleterious.  The evidence (and my own personal experience doing mutagenesis) suggests that the majority of amino acid mutations are neutral.  This is why homologous proteins can share less than 60% sequence identity, yet have the same structure and function.

But mutational tolerance is undoubtedly context-dependent, and probably varies greatly from protein to protein.  I would expect that proteins that interact with lots of other molecules, like p53 or calmodulin, would tolerate far fewer mutations than a simple enzyme that gets overexpressed.

Comment #31887

Posted by Longhorm on May 24, 2005 05:18 PM (e) (s)

Dembski claims that specified complexity is evidence for ID.  He claims that when one sees SC, then one should infer design, because SC objects cannot come about naturally.  (If the meaning of “natural” isn’t quite clear, that’s Dembski’s problem — he is the one drawing the dichotomy between “natural” and “intelligent” causes, and claiming to find evidence of the latter by way of rejection of the former.)  Therefore, Dembski is obligated to demonstrate that SC objects cannot evolve naturally.  Otherwise, his argument fails.

What do you mean “naturally?”  I don’t think he is obligated to demonstrate that “SC objects cannot evolve naturally?”  It is sometimes hard to show that there is not a causal linkage.  So I think he is obligated to tells us some of the event(s) he believes were caused by the designer(s).  That helps us determine if his claim is reasonable.

It would be so difficult for him to show that “SC objects cannot evolve naturally.”  I mean maybe he should try anyway.  I wouldn’t object to him doing so.  I just don’t think he is obligated to.  I don’t even know where to begin.

What he should do is just say what event(s) the designer caused.  Did the designer turn dust directly into an elephant?  That is the simple positive claim.  And then we could all go home. 

And it’s not anyone else’s obligation to demonstrate that any given thing can evolve naturally.

I strongly agree. 

But aside from whether the event was “natural” or not, the question I’m interested in is whether the event occurred.  Given how I think you are using the phrase “natural,” it is logically possible that non-natural events occurred.  The question is: Did the ones Dembski thinks occurred actually occur?

Sometimes people offer claims that should not be dismissed as “non-science.”  It is often clearer just to assess they claim, and determine whether the alleged event actually occurred.  Sometimes we won’t know for sure whether it occurred.  However, sometimes we are able to determine, with at least a high degree of probability, that an alleged event did not occur.  For instance, some people think the universe is about 6,000 years old. 

And the data Dembski presents is specified complexity.  (Which of course is not data, but an argument that certain sets of data must be the result of ID.)  For a long time, he claimed that “the math” showed that specified complexity could not evolve.  Skeptics asked to see these calculations that suposedly demonstrated that SC objects could not come about by natural means.  Dembski vacillated and kept promising that they were forthcoming (after claiming that they had already been done).  Finally, in No Free Lunch, he presented the one (1) calculation that he has done to date.  This was on the flagellum of  E. coli, for which he made the grand discovery that a random combination of amino acids is extremely unlikely to result in a functioning flagellum.  Since no one thinks that happened anyway, the calculation was pointless.  No evolutionary scenarios were taken into account.  His excuse for this was that irreducible complexity made the evolution of the flagellum impossible, which if true, would be enough by itself to conclude that it couldn’t have evolved naturally.  Thus, Dembski’s SC doesn’t actually add anything of merit.

I don’t think Dembski’s claims — at least those that I’ve seen — add anything of merit, either.  But I’m more concerned about whether we are justified in believing that they are true.  Some of the things of I’ve seen people say are just too vague.  But some people say that the known universe is about 6,000 years old.

Comment #31888

Posted by SWSchaeffer on May 24, 2005 05:23 PM (e) (s)

Longhorn posts:

SW, thanks for the post.  How are you using the word “deleterious?”  And what evidence is that there most mutations are “deleterious?”  And do you have a reference?  Is there an especially good article on the issue?  I hear this discussion a lot, and it seems like there is not a clear consensus.

Deleterious means that most mutation within coding sequences are likely to be lethal.  One can infer the lethality of most amino acid changes by comparing the rate of change in nucleotide sites that are in synonymous sites (sites within codons that fail to change the amino acid sequence, usually third positions) versus  nonsynonymous sites (sites within codons that do change amino acid sequence, usually first and second positions).  Synonymous codons should reflect neutral sites and nonsynonymous sites may or may not be subject to selection.

If the ratio of nonsynonymous to synonymous changes for a gene is less than one then nonsynonymous substitutions are largely deleterious.  Most genes tend to fit this pattern.

If the ratio of nonsynonymous to synonymous changes for a gene is greater than one then nonsynonymous substitutions are subject to diversifying selection.  Major histocompatibility genes fit this pattern.

If the ratio of nonsynonymous to synonymous changes for a gene is equal to one then nonsynonymous substitutions are neutral.  Pseudogenes tend to fit this pattern due to the lack of selective constraints.

Comparative genome studies where all genes are compared between two species show that the majority of genes have a ratio of synonymous to nonsynonymous changes are less than one. See the open access article by Richards et al. (2005) Genome Research 15:1-18.

But not all amino acid changes are deleterious.  The evidence (and my own personal experience doing mutagenesis) suggests that the majority of amino acid mutations are neutral.  This is why homologous proteins can share less than 60% sequence identity, yet have the same structure and function.

But mutational tolerance is undoubtedly context-dependent, and probably varies greatly from protein to protein.  I would expect that proteins that interact with lots of other molecules, like p53 or calmodulin, would tolerate far fewer mutations than a simple enzyme that gets overexpressed.

Again, my view is based on the nonsynonymous to synonymous ratio.  Mutating genes in the laboratory may not reflect the acceptability of these changes in nature within an organism. 

I base my view on the filter of

Comment #31895

Posted by Steve Reuland on May 24, 2005 06:03 PM (e) (s)

Longhorn wrote:

What do you mean “naturally?” 

The same thing Dembski does.  He claims that “natural” means “non-intelligent”.  This is a highly problematic definition, but for now it will suffice.

I don’t think he is obligated to demonstrate that “SC objects cannot evolve naturally?” 

His argument is that SC objects cannot evolve naturally, which means they must have been designed.  Therefore, he is obligated to demonstrate it!  I don’t know I could make it any more plain than that.  When one makes claim X, one has to justify claim X.  Perhaps you think that Dembski is making a different claim than he really is.

It is sometimes hard to show that there is not a causal linkage.  So I think he is obligated to tells us some of the event(s) he believes were caused by the designer(s).  That helps us determine if his claim is reasonable.

I agree, but this is not how Dembski operates.  He is not putting forward any hypotheses about how this or that living thing managed to come about — he is only claiming to “detect design”.  He doesn’t know how living things were designed, but he knows they were.  This may be a bogus side-stepping of crucial issues, but one can still take him at face value and see if his methodology has really detected design like he says it has.  His methodology works like this:  If something couldn’t have evolved naturally (e.g. specified complexity), then it must have been designed. 

My point this whole time is that Dembski has not demonstrated that specified complexity, given the definition used by Orr, could not have evolved naturally.  Thus, his design detection methodology, even if we overlook the many logical flaws, has not detected design.  It would have been nice for Orr to have pointed that out.  Dembski’s argument fails quite trivially and doesn’t require any special lessons from evolutionary theory to refute.

It would be so difficult for him to show that “SC objects cannot evolve naturally.” 

Of course it’s difficult.  But that’s Dembski’s problem.  That is the claim he is making, that is what he has to demonstrate.  No one said he was being reasonable.

Comment #31901

Posted by Longhorm on May 24, 2005 06:15 PM (e) (s)

Steve posted:

And it’s not anyone else’s obligation to demonstrate that any given thing can evolve naturally…

I don’t think people should even use the word “natural.”  What does that mean?  We should just say the Broncos won two Super Bowls.  Mount St. Helens erupted.  Lincoln was president before Grant.  Some fish evolved into amphibians.  I had cereal for breakfast.  I don’t even like the phrase “natural selection.” 

This approach is clean and simple.  And the “natural”/”non-natural” distinction tends to confuse things.  It’s a little vague. 

However, let’s say someone says that God specially intervened to cause a given event, but God’s intervention was one cause.  And it was under the surface.  There is no sign of it.  What do you do with that? 

For instance, John Elway threw that touchdown pass, but God intervened to make sure it was completed.  But it was a discrete intervention that you couldn’t tell with the naked eye. 

Or suppose someone says: “Self-replicating molecules evolved into humans, but God helped the process along the way.”  With that level of clarity, it is hard to know what to do with the claim.  But it shouldn’t be part of the official curricula for public school science classes.  And I would want to pursue the issue with the person to get a better sense of what he thinks happened.  To see if we are justified in believing that it did.

Comment #31903

Posted by Steve Reuland on May 24, 2005 06:17 PM (e) (s)

SWSchaeffer wrote:

If the ratio of nonsynonymous to synonymous changes for a gene is less than one then nonsynonymous substitutions are largely deleterious. 

No, it simply means that some substitutions are deleterious, and that these have been weeded out by purifying selection.  It does not mean that all substitutions are deleterious, or even that most of them are.  For this to be the case, the nonsynonymous to synonymous ratio wouild have to be close to zero — in other words, a situation in which hardly any amino acid substitutions had occurred at all.

Comment #31904

Posted by Steve Reuland on May 24, 2005 06:21 PM (e) (s)

Longhorn wrote:

I don’t think people should even use the word “natural.” 

In case you haven’t noticed, the IDists use this term constantly.  Take it up with them.  I agree that it’s problematic, but it’s easier to address their arguments if one doesn’t first have to invent new terminology.

Comment #31906

Posted by Frank J on May 24, 2005 06:26 PM (e) (s)

Since we’re picking on Orr, I’ll add my 2c:

H. Allen Orr wrote:

Senator Rick Santorum, a Pennsylvania Republican, has argued that “intelligent design is a legitimate scientific theory that should be taught in science classes.”

Actually Santorum has backed off his demand that ID be taught, probably on the advice of the Discovery Institute. Last I heard he advocated the phony “critical analysis” which is just as bad. Also I don’t like Orr’s frequent use of the word “Darwinism.” I’m sure that he means simply “Darwinian evolution,” but most nonscientists will infer the anti-evolutionists’ connotation.

Criticisms in this thread aside, few others can match Orr at demolishing ID pseudoscience.

Comment #31909

Posted by Longhorm on May 24, 2005 06:34 PM (e) (s)

Steve posted:

Of course it’s difficult.  But that’s Dembski’s problem.  That is the claim he is making, that is what he has to demonstrate.  No one said he was being reasonable.

I think it would be fine if Dembski tried to show that “SC objects cannot evolve naturally.”  But I don’t have a problem if he doesn’t.  But, after thinking about it, it is not important to me either way. 

What is important is for him to say what events he thinks the designer caused.  That is very clear.  The claim “Humans being were designed” is just too unclear for me to determine whether I should accept the claim.  Did a designer have some role to play the existence of humans?  Well, I would have a better chance of determining that if Dembski would just say what he thinks the designer did.  I was born.  Did the designer turn dust directly into the first two humans?  No.  And if Dembski said that, then we could determine that he is mistaken. 

But he is being cagey.  He is not letting on as to what the designer did.  It’s all very secretive.  I’m not asking for the identity of the designer.  I just want to know what the designer did.

But the claim “cats were designed” should not be taught in public school science classes.  No way.  It is very vague.  And it might make some kinds think that God turned dust directly into two cats.  And that didn’t occur.

Comment #31913

Posted by Longhorm on May 24, 2005 06:40 PM (e) (s)

Steve posted:

In case you haven’t noticed, the IDists use this term constantly.  Take it up with them.  I agree that it’s problematic, but it’s easier to address their arguments if one doesn’t first have to invent new terminology.

Maybe I didn’t make my point well.  My point is that those of those who think evolution is well-supported and want it taught well in the public schools should not use the word “natural” in reference to evolution.  We should just say what we think happened.  And, if necessary, present data that we think enables us to determine this.  The “natural”/”non-natural” issue tends to confuse.  And many U.S. citizens, when they hear the word in reference to evolution, tends to make them less apt to accept evolution and learn about it.

And I don’t even know what “natural” means.  Some alleged events have occurred.  Or at least we are justified in believing that they did.  And some alleged events did not occur.  At least we are justified in believing that the didn’t.

Comment #31917

Posted by Frank J on May 24, 2005 06:42 PM (e) (s)

And now, equal time to defend Orr:

Steve Reuland wrote:

1.  Orr unfortunately falls for the old trick that the IDists accept an old Earth and some form of evolution.  This is not the case.  As I’ve pointed out here many times, the ID movement takes no position on these issues.  This crucial distinction is necessary for understanding the unscientific nature of the ID movement.

Again, I agree that ID officially takes no position other than “anything but ‘Darwinism’.” But Orr probably has in mind Behe in particular, who is quite up front about accepting an old earth and common descent. Most other IDers evade the age of the earth question and spin vague arguments against common descent. But they rarely if ever challenge Behe directly. Why? My guess is that they are smart enough to know that YEC and classic OEC are nonsense. Otherwise they’d have nothing to lose, and plenty of scientific credibility to gain by healthy internal debates.

Comment #31921

Posted by Steve Reuland on May 24, 2005 06:49 PM (e) (s)

Longhorn wrote:

I think it would be fine if Dembski tried to show that “SC objects cannot evolve naturally.”  But I don’t have a problem if he doesn’t.  But, after thinking about it, it is not important to me either way. 

Then apaprently, you don’t have a problem if he cannot demonstrate what he claims to have demonstrated.  Dembski is not merely saying that things were “designed”, he is making a specific argument to that effect.  That argument was critiqued by Orr, but he missed the most important part IMO.

I usually tell myself that two detailed explanations are all I’ll do before concluding that either I’m not communicating well, or that my audience just isn’t listening.  Therefore, I’ll stop now.

Comment #31923

Posted by Longhorm on May 24, 2005 06:53 PM (e) (s)

SWSchaeffer posed:

Deleterious means that most mutation within coding sequences are likely to be lethal.  One can infer the lethality of most amino acid changes by comparing the rate of change in nucleotide sites that are in synonymous sites (sites within codons that fail to change the amino acid sequence, usually third positions) versus  nonsynonymous sites (sites within codons that do change amino acid sequence, usually first and second positions).  Synonymous codons should reflect neutral sites and nonsynonymous sites may or may not be subject to selection.

If the ratio of nonsynonymous to synonymous changes for a gene is less than one then nonsynonymous substitutions are largely deleterious.  Most genes tend to fit this pattern.

If the ratio of nonsynonymous to synonymous changes for a gene is greater than one then nonsynonymous substitutions are subject to diversifying selection.  Major histocompatibility genes fit this pattern.

If the ratio of nonsynonymous to synonymous changes for a gene is equal to one then nonsynonymous substitutions are neutral.  Pseudogenes tend to fit this pattern due to the lack of selective constraints.

Comparative genome studies where all genes are compared between two species show that the majority of genes have a ratio of synonymous to nonsynonymous changes are less than one. See the open access article by Richards et al. (2005) Genome Research 15:1-18.

SW, thanks for the post and the link.  I don’t understand it all right now.  But maybe I will later.

But here is my question.  Let’s say John Drake is right on mutation rates.  If most mutations were “fatal” and humans are averaging 1.6 new mutations per sexual generation among coding genes, then wouldn’t humans be in trouble?  Maybe humans are the exception rather than the rule.  And most other species are different.  But look at mice.  According to Drake, mice are averaging about 1 new mutations per sexual generation among coding genes.  Wouldn’t it put mice in jeopardy if most mutations to coding genes were fatal.

And what about RNA-based lytic viruses?  According to Drake, they are averaging one new mutation per division.  If most of those mutations were fatal, it seems like it would be bad for lytic viruses. 

That said, some species average fewer mutations than the species I’ve mentioned.  For instance, C. elegans average only .036 new mutations per sexual generation among coding genes.

Comment #31933

Posted by Longhorm on May 24, 2005 07:13 PM (e) (s)

Steve:

Then apaprently, you don’t have a problem if he cannot demonstrate what he claims to have demonstrated.  Dembski is not merely saying that things were “designed”, he is making a specific argument to that effect.  That argument was critiqued by Orr, but he missed the most important part IMO.

I usually tell myself that two detailed explanations are all I’ll do before concluding that either I’m not communicating well, or that my audience just isn’t listening.  Therefore, I’ll stop now.

No, you are communicating well.  And what you say is reasonable.  I see your point.  I appreciate your taking the time. 

I guess my point is that I want to see Dembski just say what he thinks happened.  And I feel that way about a lot of the people who refer to themselves as proponents of “intelligent design.”  They just are not clear about what they think the designer did.  In contrast, the young universe creationists tend to offer nice clear claims. 

And people who think evolution happened offer clear claims, as well.  I’m talking positive claims.  Not negative claims.  For instance, I say some fish evolved into amphibians.  So the first organisms to live on earth that were very similar anatomically to today’s frogs were born by their mothers.  They were hatched. 

I think Orr left out the most important parts, too.  First, that the people who refer to themselves as proponents of intelligent design often fail to offer clear positive claims.  Second, it is overwhelmingly probable that self-replicating molecules evolved (through reproduction) into all the organisms that have lived on earth.  Finally, when the ID people do offer positive claims, the claims are often wrong, or at least totally implausible.  For instance, the other day in a debate with William Provine, Stephen Meyer said that humans and apes do not share common ancestors.

Comment #31950

Posted by 'Rev Dr' Lenny Flank on May 24, 2005 07:48 PM (e) (s)

Remember, given the usual tactics and the likely partisan nature of the audience, I don’t anticipate getting more than one good shot, at most, so please feed me that one killer question or comment you’ve always wanted to toss old Bill’s way…!

There is only question that matters.

*ahem*

What *is* this scientific theory of ID, and how do we test it using the scientific method?

Until he answers that, nothing else he says matters.

Comment #31951

Posted by Longhorm on May 24, 2005 07:51 PM (e) (s)

I posted:

But here is my question.  Let’s say John Drake is right on mutation rates.  If most mutations were “fatal” and humans are averaging 1.6 new mutations per sexual generation among coding genes, then wouldn’t humans be in trouble?  Maybe humans are the exception rather than the rule.  And most other species are different.  But look at mice.  According to Drake, mice are averaging about 1 new mutations per sexual generation among coding genes.  Wouldn’t it put mice in jeopardy if most mutations to coding genes were fatal.

I didn’t make my point clear.  Let me try again.  Maybe humans and mice average a much smaller percentage of fatal mutations than most other organisms do.  But if the percentage of fatal mutations that they averaage is similar to that of other organisms in general, it would be hard for me to make sense of the idea that, among all organisms to live on earth, most mutations to coding DNA have been “fatal.” 

I posted:

In his book The Ancestor’s Tale, Richard Dawkins deals with this issue.  He definitely seems to think that most mutations to non-coding DNA are “neutral.”  As for coding DNA, his position is harder to tease out.  He seems on the fence to some extent.  But I would say that finally he is more sympathetic with your view.  He speaks favorably of Tomoko Ohta’s position, namely that most mutations to coding DNA are “nearly neutral instead of completely neutral.”

I re-read Dawkins on the so-called “neutral theory of mutation.”  After reading it again, he seems pretty sympathetic with it.  At one point he said: “I have always had a lot of time for the so-called neutral theory associate with the reat Japanese geneticist Motoo Kimura, or its extention, the ‘nearly neutral’ theory of his collaborator Tomoko Ohta.”

Comment #31952

Posted by 'Rev Dr' Lenny Flank on May 24, 2005 07:51 PM (e) (s)

The anti-evolutionists’ backup plan has to be to undermine the definition of science itself within the curriculum.  We saw hints of this in the recent Kansas business.  Really, it’s been implicit all along in Dembski’s writing.  The only way for his “Explanatory Filter” to be anything other than an Argument from Ignorance is to change the rules of science itself: to flip the benefit of the doubt away from science, and to throw out methodological naturalism to admit supernatural explanations.  As their direct assaults on evolution lose traction with the public (as the coverage of Kansas suggests they will), we can expect more of this Plan-B philosophical fancy footwork instead.

Alas for them, any such plans to change the definition of science to incldue “supernatural”, is DOA in court.  A clearer example of using classrooms to advance religion, could not be offered.  The plan is not only already a corpse, but is already starting to stink up the place.  The burial will be a quick one.

Comment #31961

Posted by Brian C.B. on May 24, 2005 08:25 PM (e) (s)

I can’t believe the thread has gotten this long without:

“Are we not men?”

“No! We are DEVO!”

Whip it good, comrades.

Comment #31962

Posted by Sir_Toejam on May 24, 2005 08:34 PM (e) (s)

er, just in case there remains anybody who doesn’t know what Brian is referring to:

http://www.punk77.co.uk/groups/devo.htm…

Comment #31990

Posted by Steve on May 24, 2005 11:24 PM (e) (s)

Any good mathematical evolutionary biology grad student would probably give Dembski a run for his money on these topics.

Oh Hell, anybody with a decent grasp of statistics could probably do it.  Dembski is a knave and a liar when it comes to statistics.  Look at his essay where he poo-poos the Bayesian approach for being too subjective.  Neverming that the Specified part of CSI is just as subjective (if not moreso) as well as his probability bound of 10^-150.  Orr was very, very nice to Dembski.

Comment #31991

Posted by Steviepinhead on May 24, 2005 11:43 PM (e) (s)

Bill Slides By, Dang!

I’m sorry to report that I didn’t make it to Dembski’s lecture at all, even though it’s been on my calendar for a month.  I realize I didn’t miss much, but it would have been fun.  Unfortunately, I got slammed at work late in the day (another attorney who should have been doing something on behalf of my client today was out of town instead, so the task fell back into my lap), and I’m still in the office at quarter to ten PDT, instead of over at Demaray Hall on the SPU campus interrogating Bill.

Thanks for all the helpful comments anyway—and now that I’m painfully aware that this bogus “institute” is here in town, I’m sure other opportunities to kibitz and “report” will arise.  And, Bill, I definitely owe you one!

Comment #32016

Posted by Nat Whilk on May 25, 2005 09:07 AM (e) (s)

Thanks for the recommendations, Steven.  Both books look interesting.  I guess it is understandable that mathematicians and biologists would have different ideas about what constitutes “extraordinarily sophisticated” mathematics.

Comment #32023

Posted by Steve Reuland on May 25, 2005 09:43 AM (e) (s)

Longhorn wrote:

I guess my point is that I want to see Dembski just say what he thinks happened.  And I feel that way about a lot of the people who refer to themselves as proponents of “intelligent design.”  They just are not clear about what they think the designer did.  In contrast, the young universe creationists tend to offer nice clear claims. 

I whole-heartedly agree.  The main failure of ID is that its proponents have never come up with a simple explanation of what happened when and how, and then marshalled evidence to back it up.  There is no theory of ID.  All they’ve got are crude antievolution arguments, which even if true, would not support their case.  And yes, this means that the YECs are behaving in a more scientific fashion than IDists are, which is pretty sad.

However, one can critique their antievolution arguments in addition to pointing out the gaping holes in their broader reasoning.  And my only point was that Orr failed to address what was really wrong with Dembski’s particular antievolution argument, which is that it simply begs the question.  The specified complexity argument, by itself, does nothing to cast doubt on the efficacy of evolution. 

Orr also should have pointed out that antievolution arguments, by themselves, don’t make a coherent theory, and that ID cannot be taken seriously as science until such a time as its proponents stop taking errant pot-shots at evolution and start trying to explain what actually happened.  In many respects, I think he was being far too generous.

Comment #32052

Posted by snaxalotl on May 25, 2005 11:44 AM (e) (s)

…obligation…

creationism has been backed against a wall with the overwhelming evidence for evolution, which is so good that the only escape, the last resort which allows you to cling to your contrary beliefs, is a logical disproof of evolution, at which point you know that despite the evidence the theory falls down.

The creationist on the street thinks that ID is precisely this - “there’s some guy who I’m reliably informed mathematically disproved evolution, even though I don’t understand the details”. Although the top ID proponents are considerably less forthright when arguing with people who have an actual clue, most supporters think ID provides an ironclad disproof, and the ID management are more than happy to propagate this belief.

Given that this is the product which is in fact being sold to the public, then yes there is an obligation to prove that there are things which can’t evolve.

Comment #32055

Posted by Longhorn on May 25, 2005 12:00 PM (e) (s)

Steve wrote:

I whole-heartedly agree.  The main failure of ID is that its proponents have never come up with a simple explanation of what happened when and how, and then marshalled evidence to back it up.  There is no theory of ID.  All they’ve got are crude antievolution arguments, which even if true, would not support their case.  And yes, this means that the YECs are behaving in a more scientific fashion than IDists are, which is pretty sad.

However, one can critique their antievolution arguments in addition to pointing out the gaping holes in their broader reasoning.  And my only point was that Orr failed to address what was really wrong with Dembski’s particular antievolution argument, which is that it simply begs the question.  The specified complexity argument, by itself, does nothing to cast doubt on the efficacy of evolution. 

Orr also should have pointed out that antievolution arguments, by themselves, don’t make a coherent theory, and that ID cannot be taken seriously as science until such a time as its proponents stop taking errant pot-shots at evolution and start trying to explain what actually happened.  In many respects, I think he was being far too generous.

I’ve been thinking more about your point, and I think you are right.  If Dembski is going to say that self-replicating molecules did not evolve into elephants, he should tell us why he thinks that.  More importantly, he should tell us what he thinks happened instead.  Either humans and gorillas share common ancestors or they don’t.  There is no third option. 

We agree that elephants exist.  Unfortunately, their numbers are shrinking because of hunting and other factors, which is deeply immoral.  But we still have some elephants.  What event proximately caused the existence of the first elephant?

Either it was born or it wasn’t.  I don’t seen any other options.  And if wasn’t born, does that mean that matter instantaneously transformed into an elephant?  Look, we have an event — the existence of the first elephant.  Some event caused it?  Which event?

I think the first organism that was fairly similar anatomically to modern elephants was born by its mother.  What does Dembski think?  In fact, I think self-replicating molecules evolved (through reproduction) into all the organisms that have lived on earth. 

Some people believe that a deity turned inert matter or nothingness — zap! — directly into an elephant.  I think that is absurd.  But at least it is a clear claim. 

Now there is one third option.  The first elephant was born, but God had a hand in that.  So Elway threw the touchdown pass, but God helped him complete it.  Well, what evidence is there for that?  What reason is there to believe that?  When I was born did God cause me to born?  Did God intervene and specifically cause my birth?  Well, what evidence is there for that?  Now the person who offers that hypothesis has the burden of proof. 

The key issue is we have elephants.  And some events caused the existence of the first elephant.  Hundreds of thousands — maybe millions — of elephants have lived on earth.  And hundreds of thousands — maybe millions — of elephants were born by their mothers.

The person who thinks that the first organism similar to today’s elephants was <i>not</i> born should tell us what even(s) proximately caused it to be here.

Comment #32060

Posted by Henry J on May 25, 2005 12:30 PM (e) (s)

Re “then yes there is an obligation to prove that there are things which can’t evolve.”

Well, cars and planes can’t evolve, and those are “things”. ;)

Comment #32062

Posted by Henry J on May 25, 2005 12:32 PM (e) (s)

Re “Humans average 1.6 new mutations per sexual generation among coding genes;”

Does that include mutations that cause miscarriages?

Henry

Comment #32073

Posted by Longhorn on May 25, 2005 01:29 PM (e) (s)

Does that include mutations that cause miscarriages?

I got the number from an article by John Drake.  He is at the Laboratory of Molecular Genetics in North Carolina.  I think Drake was referring to all mutations.  He uses the expression “spontaneous mutation.”  But he never says how he is using the expression.  It’s my impression that the expression “spontaneous mutation” is sometimes used to distinguish between, on the one hand, mutations caused by events such as getting hit with radiation and, on  the other hand, mutations that don’t have a clear causal agent.  I don’t like the expression “spontaneous mutation,” because it gives the impression to some people that some mutations are uncaused events.  And that makes no sense.

The important thing is that cells divide.  Some cell-divisions are followed by the daughter-cell having a different genome than the genome of its parent-cell.  There are different times a cell can come into being with a new mutation.  First, in the creation of sex cells.  Second, after a sperm-cell has fertilized an egg-cell and before the organism is conceived.  Third, if the organism is just living.  When humans get cancer, those are mutated cells. 

An important issue for study is to learn more about the kinds of events that cause cells to come into being with new mutations.  It is interesting that some kinds of organisms mutate more frequently than do some others.  And some parts of the genome tend to mutate more frequently than other parts.  Scientists call these “hot spots.”  For these reasons and others, there is excellent reason to believe that some organisms having produced the number of offspring they did contributed significantly to the occurrence of many mutations. 

Here is a link to the abstract to Drake’s article: 

http://www.annalsnyas.org/cgi/content/abstract/870/1/100…

Unfortunately, in the abstract, he doesn’t give the numbers on humans, mice and some other organisms.

Comment #32111

Posted by Salvador T. Cordova on May 25, 2005 02:57 PM (e) (s)

A very good refutation of Orr’s article can be found here:

http://www.ideacenter.org/contentmgr/showdetails.php/id/1345…

Comment #32114

Posted by frank schmidt on May 25, 2005 03:03 PM (e) (s)

Salvador is wrong on both counts: it is neither good nor a refutation.

Comment #32119

Posted by Sir_Toejam on May 25, 2005 03:11 PM (e) (s)

“Salvador is wrong on both counts”

what else is new?

Comment #32139

Posted by Tanmoy Bhattacharya on May 25, 2005 04:02 PM (e) (s)

Re discussion about neutrality.

I have never posted to a blog before: so am unclear on etiquette. Please ask for references if you want me to backup anything other than my own opinions in what I write below.  For my opinions, I am only willing to argue.

First, I think most geneticists wont define neutral as the antonym of lethal: to many, it merely indicates that the number of descendants of the mutant are less in number, on average and after many generations, than the number of descendants of the non-mutant, for whatever reasons.

Second, two mutations, each of which is deleterious, can together be advantageous/neutral.  Covariation is surprisingly common: just looking at HIV, an organism whose phylogeny I study, and looking at a random region in the genome, I can see by eye clusters of say tens of codons in the genome which seem to be changing together more often (I mean the `togetherness’ is more often) than I would expect by chance.  Of course, this may be because my current model of change is too approximate; and I am actually doing it more formally since I need to control this covariation aspect for something else I am doing. The problems observed in offsprings of parents showing hybrid vigour could be due to similar effects, but of reassortment of covarying past mutations.

Third, the advantage or fitness depends on the environment.  There  is evidence that when HIV enters a new population of humans, the mutations which were worse in a previous population now fare better. (So if you want to lead a risky life, you may be better off somewhere where everyone else looks different!) The environment can also be shaped by the mutations themselves: the evidences for the rare sympatric speciations can be used as examples.  For reasons related to this, I do not like using HLAs in this kind of argument.

Fourth, as Eigen stated, a rate of about one mutation/ unit/ generation (where, now we may argue that unit is most often the coding sequence in an individual, or maybe the segregating units) may be optimal in evolution if most mutations are slightly harmful.  We do observe this in nature, of course with a large error of about a factor of 100 (DNA viruses seem to be exceptional, but that is probably a side effect of them not being independent replicators), but given that the mutation rates per site in the genome varies over a range of about a million, this does look like extraordinary tuning of the genome size.

Empirically, even though random mutations do maintain a high level of function of the protein, the mutated protein often does not often work *exactly* the same. (Remember that the probability of getting an exact value one for a random number uniformly distributed between, say, half and two is exactly zero.) Since growth of living forms by reproduction is an exponential in time process, even small differences show up in populations.  In practice, to take a random example, out of 1003 random positions in a not too variable poly protein I have open in front of me, I have 750 positions where it is the same amino acid in almost all (techincally, greater than 97%) the HIV sequences I am looking at, and the rest show only two possible amino acids (almost).

On the other hand, there is enough reason to believe `almost neutrality’: differences are often slight.  One is not asking the question whether something is `precisely equal’, rather whether the difference is small enough compared to the rate at which mutations at related sites (which can turn this into an advantage) appear, and with respect to the time over which the environment changes enough to turn deleterious into advantageous.  Also, this `almost neutrality’ is definitely not the expectation for all mutations where an organism finds itself in a novel environment (like HIV when it first sees a protease inhibitor, or sees a new population of humans with a different HLA distribution).

Comment #32141

Posted by Tanmoy Bhattacharya on May 25, 2005 04:04 PM (e) (s)

Oops, found a typo after posting, sorry.

`to many, it merely indicates’ —> `to many, deleterious merely indicates’.

Comment #32145

Posted by Longhorn on May 25, 2005 04:16 PM (e) (s)

Tanmoy, thanks so much for the thoughtful, interesting post.

Comment #32147

Posted by Ed Darrell on May 25, 2005 04:26 PM (e)