As Neil Rickert was foolish enough to grant me posting rights, I had better take advantage of the offer before the sysops sensibly change their minds.
Consider an argument used consistently by the ID community: that the natural processes of genetic mutation and environmental selection acting upon the resultant variation are incapable of generating speciation.
Their recurring metaphor is of improbable islands of fitness separated by unbridgeable seas of non-functionality. Even if the hill-climbing capability of “random mutation plus natural selection” is real (and even incorporating unselected allele frequency changes), evolution can’t work because “you can’t get there from here”. For brevity, let me acronymise this argument as CANTSWIM.
CANTSWIM has a great many shortcomings as a metaphor for how evolutionary processes actually work, and you don’t need me to enumerate them. But let me hand over the title deeds to the evolutionary farm, and assume that CANTSWIM is factual.
At the same time, the ID community has adopted the claim made by the leadership of the ENCODE program that >80 per cent of the human genome is functional. The devil, of course, is in the detail of how “functional” is defined.
Let me leave aside whether ENCODE’s leadership or the ID people are using the term appropriately. In my opinion, if insertion of a random nucleotide sequence into a genome yields a bare biochemical response but no downstream effect on the phenotype, then that is prima facie evidence that the use of the term “functional” in this context is trivial.
However, let me go further and grant the whole house of functionality to the ID proponents. Let me grant that the vast majority of base sequences in the human genome are functional in the sense of directly determining the phenotype of an individual human being via protein expression or regulatory control. Let me dub this argument as CANTCURL.
Now, it seems to me that CANTCURL and CANTSWIM contradict each other.
If CANTCURL is true, then CANTSWIM must be invalid. That is to say, CANTCURL demands that virtually any change in a genome results in new functionality (by definition, since if they form part of the >80 per cent, new sequences must be functional, and at the same time “not previously present in the genome”).
Surely it follows that it is easy, simple, inevitable, a matter of natural course, to get to some new functional island from whatever atoll the organism happens to be sitting upon. All it takes is to wait a while and new functionality will arise, because CANTCURL guarantees that virtually any genomic configuration is functional. To wax metaphorical, the CANTSWIM sea must be shallow indeed.
Conversely, if CANTSWIM is true, then CANTCURL must be invalid. If the islands of functional phenotypes are truly separated by unbridgeable abyssals of impossible configurations, then how can functionality be present in virtually any nucleotide sequence?
I do not believe that the contradiction between CANTCURL and CANTSWIM is a purely one of logic or of definitional word games. Resolving the apparent contradiction is susceptible to experiment.
Certainly the truth value of CANTCURL can be tested by experiment (the mouse genomic excision experiment, for instance). Further experimentation on specific genome sequences should reveal whether their biochemical reactivity actually maps onto regulatory or protein expression, or is simply genomic noise.
CANTSWIM is more difficult to examine experimentally, because its underlying assumptions about the structure and temporal stability of “fitness landscapes” are so poorly defined.
For example, the overall “fitness landscape” of an organism must surely be some vastly multidimensional aggregation of simpler fitness “landscapes” derivable for every individual genetic locus that is expressed in its phenotype.
How does one aggregate the antelope fitness landscape for escape speed with the fitness landscape for resistance to tsetse fly with the fitness landscape for digestion efficiency?
Even so, what we do know is that experiments such as Lenski’s longitudinal series demonstrate that CANTSWIM cannot possibly be universally true.
My purpose in posting this is to invite comment on whether this apparent contradiction in the ID argumentation is real, or am I simply misunderstanding what they are claiming. All comments and abuse welcome.
Apologies to the Painted Jaguar.
Thanks for posting. It’s good to have occasional new topics, to keep the discussion fresh.
I’m not a biologist, so I’m not completely sure about the ENCODE results. The impression I am getting, based on what is said by biologists that I respect, is that the ENCODE results are being over-hyped.
The reaction of the ID community to ENCODE and, before that, to junk-DNA, has always seemed odd. It seems to be driven purely by theology. Of course, all of ID is driven by theology, but at least there’s a pretense of doing science. It’s been hard to see where that pretense fits in with junk DNA. I suspect that they are deeply offended by the “junk” part of that name, and think it an insult to their imagined designer.
I’m not a biologist either, at least not lately, but I don’t think that the ENCODE program’s problem is fundamentally one of hype – Ewan Birney was very explicit about why he decided to apply the term “functional” to results that indicated mere biochemical action. The broad sweep of ENCODE results seem to me to be amazingly interesting, just not supporting the publicity tag applied to them by the program head(s).
Be that as it may, the purpose of my post is to take a look at yet another case where the Big Tent of ID seems to be embracing two utterly incompatible concepts. As far as I can see, the two arguments can’t be right at the same time (but both of them can be wrong at the same time).
The isolated islands metaphor has been eroding steadily. The only thing left of ID is irreducible complexity, and that is based entirely on the premise that whatever is was meant to be.
timothya,
That would be true only if new sequences were almost always retained in the genome. An IDer could argue that most new sequences are discarded by purifying selection, and that the ones that are retained are designed sequences that confer an advantage on the organism and are thus favored by selection.
In other words, even if 80% of the retained sequences in the genome are functional (in the CANTCURL sense), that doesn’t mean that 80% of possible new sequences are functional. You need to account for the percentage of new sequences that are actually retained.
And how dare you post outside of UD without KF’s permission!
KF cracks me up.
It is best that timothya shake the dust of UD from his feet. How, in all conscience, can one contribute to a forum in which one of the most prolific posters spells “inoculations” as “innocculations”?
Neil Rickert wrote:
and timothya wrote:
(Leaving aside the relation of all this to ID)
I would say that the ENCODE results are a highly useful and massive amount of data on gene control regions. In that sense, no hype, very important science, well worth the effort. But yes, “not supporting the publicity tag applied to them by the program head(s)”. And the problem there is that Ewan Birney and his co-spokespeople were deliberately trying to prod the popular science press into a major reaction. And they succeeded brilliantly, getting multiple, widely-noticed reports, with almost all of them drawing the dramatic lesson that Science Has Now Shown That There Is No Junk DNA.
The problem is that this is not what Science Has Shown. Birney was engaging in deliberate hype to get a blaze of publicity. In the process he put us in the situation where the public has now concluded that the idea that most DNA is junk is wrong (and was always wrong). But it isn’t — there is still good reason for thinking that most DNA is junk. It is probably going to take us a decade of argument to persuade the public that junk DNA has not gone away. A very sad situation.
So the scientific work of ENCODE was not at all hype. But the hype in the way it was presented created a massive false impression, and that is why so many molecular evolutionists are so mad at Ewan Birney.
But the point of the OP is that universal function and isolated islands are incompatible notions.
Particularly since it appears there can be mutations in most of that functional code without affecting viability.
The words hoist and petard come to mind.
keiths
Your point is well made. However my purpose was not to argue the toss about the scientific validity of either proposition. Either one could be right, at least in principle, both could be wrong as matters of fact, but they can’t both be right at the same time, either in principle or in fact.
timothya,
I disagree. CANTSWIM and CANTCURL are compatible in principle (though I think we agree that both of them are false in fact).
CANTCURL merely holds that most of the genome has phenotypic effects, which means that most of the sequences in the genome are viable. You’ve leapt from “most of the sequences in the genome are viable” to “most sequences are viable”, which is an unjustified extrapolation. It neglects the effects of selection.
The only sequences that persist in the genome are those that survive many rounds of purifying selection. Any nonviable sequences will be filtered out. Therefore, to say that 80% of the sequences in the genome are viable does not imply that 80% of possible sequences are viable. Far from it.
For example, even if only 0.1% of possible sequences were viable, it would still be possible to construct a genome in which 80% of the sequences were viable. You would just have to select most of the sequences from the 0.1% subset of the overall sequence space.
I don’t think most of that 80 percent is conserved.
petrushka,
True, but timothya grants it arguendo as part of CANTCURL:
I think the point is that if it is functional and not conserved, than the percentage of sequence fragments that are functional is very high.
Shiver me timbers, I think ye poxed sons of whores be assuming that mutations are random.
All mutations are the result of His Noodly Appendage.
R’amen, brothers.
First of all, the ENCODE results were preliminary and studied only some of all the cell types in the body. Therefore, it should be expected that many more areas of the genome that used to be referred to as “junk” will be seen to be functional in certain cell types not yet studied. Therefore, I think we’ll find that the epigenetic (regulatory) areas of the genome will be, if anything, even more complex than we know so far.
Secondly, that doesn’t mean that some parts of the genome are not junk. We already know from gene mapping that there are parts of the DNA that are remnants of viruses that are inoperative and yet have been dutifully copied and passed on because they have neither a positive nor a negative influence on fitness. This idea that the genome has a capability of “purifying” itself is misguided. Genetic drift is known to cause mutations that are neutral and which form variation which may, in the future, provide some advantage if the environment changes.
Thirdly, it has been known for some decades that there is such a phenomenon as “directed” or “adaptive” mutations. Although there is a lot of controversy as to how it works, it is known that adaptive mutations can occur in response to a new environment at a much faster rate and in a more precise way, focused on the particular gene(s) needed to survive, than by mere randomness. Physicists, and even Crick, have even proposed that quantum mechanics is involved. For a fuller discussion, please see here.
Adaptive Mutations and Quantum Mechanics
There are more things in heaven and earth, Horatio, Than are dreamt of in your philosophy.
DNA_Jock,
Au contraire. Design is the glue that holds CANTSWIM and CANTCURL together.
CANTSWIM says we are stuck on meatballs of functionality, surrounded by the sauce of non-function. The sauce is unbridgeable by random mutation plus natural selection, but a rameniscient, ramenipotent deity can pluck functional sequences from scattered meatballs and combine them into a genome full of functionality, as required by CANTCURL.
A bit of parmesan seals the deal.
billmaz
This a little OT, but . . .
Apropos directed mutations, [t]here is a small but crucial distinction between somatic cells and germline cells. I know of no established examples of proposed adaptive mutations involving the latter. In other words, even if it exists, the phenomenon has zero effect on the evolutionary history of any organism (excepting the second degree case that a capability for hypermutagenesis in immune systems is likely to be adaptive).
timothya,
The rest of us can’t see billmaz’s comment because it’s in moderation, probably because he is a new commenter.
As the owner of the thread, you can click on ‘Allow user to comment without moderation’ and his comments will be visible to the rest of us.
keiths
We both agree that CANTSWIM and CANTCURL are likely wrong as matters of fact. But, as you say, my argument is that they are incompatible as a duality.
Your responses indicate two ways of making them compatible: godpoofery of functional sequences, or natural selection filtering out non-functional sequences.
I must admit that I can’t think of a way to resurrect (!) the incompatibility, so I must concede. Scratch one dopey argument.
timothya,
Kudos to you for that forthright acknowledgment.
billmaz’s comment is now visible. Sorry for not noticing (and fixing) that earlier.
keiths:
Leave aside the finger-waving, pompous windbaggery, and the gratuitous torture of innocent syntax, and his penchant for self-referential acronyms.
What will condemn him to the Uttermost Pit is his constant invention of rebarbative neologisms.
Bugger, did I just commit one myself?
What’s new and different about this OP?
Water does not exist, or if water does exist, life can cross it no matter the distance to the next bit of land, is no new claim.
But what’s the evidence?
Surely the claim is not limited to ID proponents.
Another IDiot argument goes down in flames.
Mung,
It’s noted that you only demand evidence for non-ID claims. You are happy to accept (as your “Why Mung is an ID supporter” thread amply demonstrated) any pro-ID claim without any evidence at all.
Mung,
You miss the sciencey posts at UD too, don’t you? 🙂
T,
Yes. But I do enjoy the philosophy.
After all, why do we do science?
OMTWO:
And your evidence is?
Why do we do science, Toronto?
You have no answer. Really?
Sorry to resurrect this thread but andyjones on UD:
has demonstrated keiths’ prediction: that IDists insist on godpoofery to reconcile the contradictions between the high-function and low-evolution postulates.
Purpose, apparently, is an essential precondition to all, any, some biological events.
By the way, if you haven’t got to it yet, Graur et. al. on the ENCODE events is definitely worth reading:
http://gbe.oxfordjournals.org/content/early/2013/02/20/gbe.evt028.short
Hee hee! He doesn’t do things by halves, does he? One wonders how population genetics and neo-Darwinism coped prior to 1972 and Ohno.
Andy Jones does rather seem to miss the point of Graur et al‘s paper. Evolution (little-e evolution, the kind no-one is supposed to have any problem with) acts upon all loci. Favoured loci are not lead-shielded and immune to polymerase error. There is constant attrition, and one does not need to adhere to some mythical ‘cult of Darwin’ to accept that as fact.
In the face of this attrition, if selection is weak or absent, there is nothing apparent with the power to stop it becoming an ex-function. So if your ‘function’ makes no difference to the organism’s survival or reproduction, it’s hard to see how one can call it a function at all, even if it pops up in an assay for (say) transcription factor binding. Jones rolls out the stock engineering analogies – but if I stuck 1000 lines of code in a program that all said “ooga = booga”, they have a ‘function’ of sorts – it is recognised by the compiler, which generates machine code, so if that was how ‘functional’ code was measured, it would pop up … but it has no role in the task. I may well have had a purpose putting them there – maybe I get paid by the line – but in a milieu that frequently copied the program with occasional random errors, changes in those lines would generate fewer ‘bad copies’ than in more significant lines. And that’s where evolution comes in. How do you keep it out?
If, alternatively, we are expected to simply ignore all the processes of little-e evolution – mutation, selection, recombination, drift – what on earth are we left with? Genes that retain their nonessential function in perpetuity, regardless what happens to them? How does that work? Alleles that are never lost from a population, whatever their effect? How come? Alleles that give an advantage but never increase? What stops them?
Graur et al. were pretty specific about the computational shortcomings of the ENCODE methodology. Presumably this would mean that if the raw measurements in each category of investigation were re-run with more realistic boundary conditions, then the results would also provide a more realistic estimate of true functionality in the genome.
That is not to say that such a recalculation is feasible (presumably the computational effort was enormous in the first place).
Your own writing. I have just made a post on a different thread that notes that you are taking part in a thread in UD where the claim is made that ID explains Large Morphological Gaps, Sudden Appearance, Convergence and Lack of clear phylogeny.
You can find that post here: http://theskepticalzone.com/wp/?p=1559&cpage=1#comment-19660
Those claims are made on a thread you are taking part in on UD and I have not seen any demands from your for those explanations. Therefore my claim is supported by your actions (or rather, lack of action) quite clearly.
Some pro ID claims are made. You stay silent, despite no evidence being offered.
Case closed.