Up Above
by MikeGeneThis entry was posted on Sunday, January 20th, 2008 at 12:30 am and is filed under The Rabbit. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site. The trackback link is: http://telicthoughts.com/up-above/trackback/

























January 20th, 2008 at 1:20 am
http://www.sciencedaily.com/re...
any thoughts?
Comment by Lord Timothy — January 20, 2008 @ 1:20 am
January 20th, 2008 at 12:39 pm
Lord Timothy:
Yeah. This is some serious historical revisionism. Now that both Charlie's version and its "neo" manifestation have been demonstrated wrong. Are we to pretend ignorance of Random Mutation as the inheritance half of Neodarwinism's RM-NS equation?
From your Nov. 19, 2007 link:
[emphasis mine] Now, why do you suppose these multinational researchers were inclined in November to surmise that worm vulvas "should" evolve in a stochastic or 'random' fashion? Could it be because the Neodarwinian explanation of inheritance (variation) was labeled 'random'?
Conclusion from that report:
Over in the sidebar under "related stories" is one dated January 20, 2008 - that's today, for those of you keeping track. Its deliberately deceptive headline reads -
New Findings Confirm Darwin's Theory: Evolution Not Random
Whaaaaaa??? Since when did "Darwin's Theory" claim that the inheritance end of evolution wasn't random? This article - a follow-up to the article above - begins…
What "opposing theory" to Darwinism/Neodarwinism is it that claimed variation was random? Can't be ID or Creationism, neither of which claim inheritance is random.
Kind of amazing that they'd try to pull this historical revisionism to pretend biologists always knew inherited traits weren't random. Looks fairly desperate to me!
Comment by Joy — January 20, 2008 @ 12:39 pm
January 20th, 2008 at 1:23 pm
Generational inheritance is observed to be decidedly non-random. We observe that parents strongly resemble their parents. Note that in the article "deterministic inheritance" is in scare-quotes indicating that the term is being used in a technical fashion. Darwin posited that most evolutionary change is deterministic adaptation. Modern researchers have found that a lot of molecular evolution may be stochastic.
It couldn't be ID or Creationism as they don't constitute valid scientific theories. The study concerns whether evolution (change over time) is primarily due to stochastic processes or deterministic processes.
If we could observe the evolution of a giraffe's neck, stochastic would mean that the neck would lengthen or shorten in a random walk over time, while deterministic would mean that the neck would tend to lengthen continuously.
The authors of the nematode study conclude, "Surprisingly, only two characters showed unbiased evolution. Evolution of all other characters was biased."
Comment by Zachriel — January 20, 2008 @ 1:23 pm
January 20th, 2008 at 2:21 pm
Zach:
Then, pray tell, what is the opposing "scientific theory" this research puts the lie to? What theory has been reduced to a catchy formula even disinterested high school students can regurgitate on demand? You know, the RM-NS thing - Random Mutation determined by Natural Selection.
But where's the bias? The story says…
"Randomly inherited and not necessarily advantageous" = RM: the 'random with regards to fitness' tenet in the RM-NS 'central dogma' of Neodarwinism.
I was taught RM-NS in school. My kids were taught RM-NS in school. You were probably taught RM-NS in school too, my grandchildren are being taught RM-NS.
We were taught that selection is deterministic, inheritance random. The conclusion in this study is selection is biased and/or there are some selection-independent constraints in operation. Talk about an inconclusive conclusion! If the constraints are selection-independent, which half of the variation-selection divide must they fall into?
Still, the Current Biology link makes it clear this study was inconclusive. It established an evolutionary bias. It doesn't say which side of the equation the bias comes from, even though the press does emphasize that this inconclusive conclusion belies the "opposing" theory that maintains variation is random with regard to fitness. Hmmm….
Comment by Joy — January 20, 2008 @ 2:21 pm
January 20th, 2008 at 2:47 pm
Thought provoking study apparently. But maybe not all that surprising if you think about it. Random mutation of the genes underlying a metric (such as developmental) trait Y will almost certainly lead to biased mutation in the trait itself, in the sense that random mutations are more likely to cause an increase (or decrease) in Y rather than vice-versa. And the reason is the non-linearity of the genotype-phenotype map. For example, take the non-linear function f(x)=x^2 (x>0), then a random change in x will on average lead to an increase in Y=f(x). (the average of f(x-delta) and f(x+delta) is larger than f(x) for constant delta.)
If there is an "optimal" vulva phenotype, whose development depends on multiple traits, then biased mutation in one of the traits will cause directional selection in some other trait(s) in order to keep the vulva phenotype invariant.
Make sense?
Comment by Raevmo — January 20, 2008 @ 2:47 pm
January 20th, 2008 at 3:45 pm
Raevmo, they're talking a non-random genetic bias. That's the randomness the un-named "opposing theory" asserts, which this study falsifies. A gene that affects the size of vertebrae or discs (to make the neck measurably longer) is not predominant in a population unless isolation allows genetic drift (the population is thus closely related to the precursor, whose variation was random in the original population). The conclusion is that drift played no role, that there was a bias in the type of variants presented in the whole population - directional toward longer necks (using the giraffe extrapolation).
Reluctance to rule out selection bias on the obvious criteria is odd, but maybe not so odd in light of reluctance to name the "opposing theory" that maintained variations are random with regard to fitness. They want to pretend current theory is is not neo-Darwinism so that this nail in its coffin won't stand out.
But when the headline reads:
New Findings Confirm Darwin's Theory: Evolution Not Random
…somebody's being less than honest. Both Darwinism and Neodarwinism rely/relied on randomness on the variation end.
I see nothing at all in these reports and abstracts to suggest that Darwin's "pangenesis" theory of inheritance has been confirmed.
I see neo-Darwinism's central dogma of random wrt fitness variation being directly challenged, NOT confirmed.
Who's theory of evolution is being confirmed by finding that evolution's not random?
Comment by Joy — January 20, 2008 @ 3:45 pm
January 20th, 2008 at 4:14 pm
Joy:
I don't think so. I've had a quick look at the original paper, and it seems to me the study concerns phenotypic evolution of developmental traits. I don't think they know the genetic basis of the traits they're studying.
No, I think you're wrong. This "opposing theory" is simply neutral evolution, which is rejected in favor of some unknown deterministic process, perhaps some mix of selection, biased mutation, developmental constraints, what have you.
I agree that the headline is a bit strong. But the results are certainly consistent with an important role for selection. Nothing indicates that the underlying genetic mutations were non-random.
Oh, please. A ridiculous strawman.
Then I suggest you look carefully again and modify your view accordingly.
Comment by Raevmo — January 20, 2008 @ 4:14 pm
January 20th, 2008 at 4:49 pm
The question is whether evolution is primarily stochastic or primarily deterministic (due to selection or constraint). In our example, did the giraffe neck meander in length and just happened to end up at its present length, or did it lengthen consistently over time, indicating selection?
They're talking about over evolutionary time scales, not generation to generation. Think about the example again.
Modern evolutionary theory posits that mutations are random with respect to fitness. But when you stand back at a distance of time, this genetic randomness may not be readily visible. The study is not concerned with the mutations in each generation, but with evolution over longer time scales. Do traits drift, or are they being forced (biased) in one direction or another.
Over the long run, does the giraffe's neck look like it is being stretched out by deterministic processes, or does it meander in length and just happened to end up in its current configuration? This seems obvious in the case of the giraffe, but not so for other organisms and other traits. And scientists are not satisfied with just what seems obvious. This study provides evidence that deterministic processes are more important, just as originally proposed by Darwin.
Comment by Zachriel — January 20, 2008 @ 4:49 pm
January 20th, 2008 at 5:11 pm
Raevmo:
Huh. Then why do you suppose they call it "deterministic inheritance"?
Neutral evolution? That doesn't sound much like "deterministic inheritance." Are you sure you've parsed the words correctly?
"Some mix of selection, biased mutation and developmental constraints" (epigenetics) sounds a lot like EAM. It doesn't sound at all like Neodarwinism.
"An important role for selection" is not news, Raevmo.
No, it's tongue in cheek satire. Darwin's pangenesis lost to Weismann within a decade.
I have looked carefully. "Deterministic inheritance" alludes quite specifically to the inheritance side of the equation. "Not influenced by drift" means selection is downgraded for this directionality.
Which evolutionary theory is it that has insisted that inheritance - variation - is "random with regard to fitness?"
Comment by Joy — January 20, 2008 @ 5:11 pm
January 20th, 2008 at 5:53 pm
It refers to "evolutionary patterns of change". Read it as "Deterministic Evolutionary History" vs. "Stochastic Evolutionary History". The authors of the scientific paper do not use the term.
"Stochastic or deterministic hypotheses." That's what is being tested. It is not testing for random mutations, which are strongly supported by other studies, but the causes of "evolutionary patterns of change".
Comment by Zachriel — January 20, 2008 @ 5:53 pm
January 20th, 2008 at 6:04 pm
Joy:
Dunno. I've never heard of the term before. It's not used in the original paper either.
I think so. The paper shows that the developmental traits have not evolved according to some random drift (neutral evolution) scenario, which might have been expected (rather naively perhaps, but not unreasonable as null-hypothesis) given that the end result of development remained constant. The results show evolution was non-random (directional), so the drift scenario is rejected. The journalists like to call that "deterministic inheritance" apparently, but I wouldn't get too hung up about that phrase.
I don't know what EAM is (do you have a link?) Maybe it doesn't sound like Neodarwinism to you because you have a somewhat cartoonish view of what Neodarwinism is.
Full disclosure: I'm not the editor of ScienceDaily. I don't decide what's news, but I think the results of the original work are newsworthy. There is still a lot of debate about the relative importance of natural selection in evolution, and this elegant paper adds to the body of evidence.
Well, sorry I took your joke too seriously.
It would seem so, but in that case it is an unfortunate misleading choice of words.
Oh? How do you figure that?
You tell me. I can't be the theory that says that natural selection is important, because selection depends on correlation between phenotypes and fitness.
Comment by Raevmo — January 20, 2008 @ 6:04 pm
January 20th, 2008 at 6:08 pm
Zach, I'm reading the two press releases from ScienceDaily and the abstract you linked. I am not a biologist, I just took some biology in school and some specialized courses pertaining to biology (crystallography and health physics) in college and worked for awhile in an application. I get my news just like other interested, not entirely ignorant but not professional lay people do. By reading the news and scanning accessible abstracts.
There are problems with the science press, and nobody seems able to figure out where they're coming from. Whether it's researchers making silly over-hyped claims, student science journalists writing spurious departmental press releases, or science press publishers writing spurious headlines and making stuff up. Or it could be a concerted effort to hype findings, re-write history, and otherwise mislead the public.
Which is it in this case? They're not talking the kind of distance you want to hold sway, nor are they talking 'founder effects' and isolation, complete with its selectively deterministic mechanism of drift. These things are specifically eliminated from consideration of this "deterministic inheritance." In 51 species of roundworms, not in giraffes.
I'd be willing to predict (from an EAM position) that the most considerable mechanism(s) in this observation will turn out to be epigenetic, rather than genetic mutation or environmental selection. Though there is genetic variation (this was noted). Once channelized, the expression suiting and timing for development adjusts and the compensations and refinements are inherited - epigentically - to operate even when the specific gene variant isn't present. Pretty soon everybody's got long necks (or workable vulvas).
No matter how they're sex-linked, the epigenetic machinery will try to (tend to) adjust for coordinated expression of influenced traits. Big heads are a male sex-linked development in humans. The ability to birth big-headed babies is female sex-linked. Humans may not have the most fool-proof reproductive system among primates, but there's 6.5 billion of us on the planet. It obviously works. Had the two opposing sex-linked developments not been adjusted for concurrently, we'd all be small-headed knuckle-walkers.
But EAM isn't Neodarwinism, it's more akin to ID. So go ahead and tell me this is all about "an important role for selection" in evolution, and tell me why that's news. I learned about selection's important role in evolution 40 years ago. If they're still spending precious research money on 'confirmation' that selection plays an important role in evolution, I've a bridge in Brooklyn I'll sell cheap… §;o)
Comment by Joy — January 20, 2008 @ 6:08 pm
January 20th, 2008 at 6:27 pm
Joy:
YAY! A conspiracy!
That's easy for you to say. Your professional reputation is not at stake. But some recent work by Craig Venter's people sheds some doubt on your prediction (Lartigue et al. 2007. Genome transplantation in bacteria: changing one species to another. Science 317). They injected naked DNA (without epigenetic markers) into cells and lo and behold:
Not exactly what your theory predicts, or is it?
Comment by Raevmo — January 20, 2008 @ 6:27 pm
January 20th, 2008 at 6:37 pm
So… you're conceding that the "deterministic inheritance" does pertain to genetic variation? Or are you still asserting that the story's about selection and inheritance has nothing to do with it?
Comment by Joy — January 20, 2008 @ 6:37 pm
January 20th, 2008 at 6:50 pm
Joy:
I asserted that?
Comment by Raevmo — January 20, 2008 @ 6:50 pm
January 20th, 2008 at 7:29 pm
The original November 2007 article:
Evolution Is Deterministic, Not Random, Biologists Conclude From Multi-species Study-
"An astonishing amount of developmental variation" that caused them to predict that it all evolved in the "expected" random fashion. Alas, not so…
"Even greater number of evolutionary changes" than they expected? Yet all were "unidirectional" toward the basic roundworm vulva with "little variation" [observable].
The conclusion was: "These results demonstrate that, even where we might expect evolution to be random, it is not."
A lot of developmental and inherited diversity, all channeled toward the same result no matter what "evolutionary changes" happened. Selection obviously had no problems with divergent development programs or genetic variations - it didn't select them out, did it? So how did selection determine what the end result would be, despite all those astonishing variations? Is this non-random selection (duh) or nonrandom variation?
Comment by Joy — January 20, 2008 @ 7:29 pm
January 20th, 2008 at 7:44 pm
Of course it's about selection and inheritance. Did you understand the example?
Each generation of giraffes varies somewhat from the previous generation some of which is due to random genetic variation. But these changes between generations are typically very small. However, over long periods of time, there may be profound morphological changes.
At the end of such a long period of time, we might see the giraffe necks are longer. The question being considered concerns change over that long interval. Do neck lengths vary stochastically and the result just happens to be a longer neck or do neck lengths become consistently longer meaning they are being pushed.
Do you understand what a random walk is?
Comment by Zachriel — January 20, 2008 @ 7:44 pm
January 20th, 2008 at 8:34 pm
Joy:
Let's try a modeling approach. Suppose the vulva phenotype is V, and it depends on developmental traits X and Y. It's quite possible that many (X,Y) pairs give rise to the same V. If selection favors a specific value of V, say V*, then selection will remove (X,Y) pairs that do not generate V*, but will keep the set of (X,Y) pairs that do generate V*. For example, say that V=aX+bY for fixed constants a and b. Then selection will keep X and Y on the line Y=(V*-aX)/b. If there's any biased mutation on X or Y (and I explained before that this is very likely), then there will be directional evolution of X and Y along that line.
Comment by Raevmo — January 20, 2008 @ 8:34 pm
January 20th, 2008 at 10:26 pm
Zach:
The research isn't about giraffe necks, Zach. It's about round worms and their vulvas. The variations and evolutionary changes were very large - "astounding" in number and kind - NOT very small at all. The outcome was uniform anyway.
Yes, I understand what a "random walk" is. That's the "opposing theory" that was soundly falsified by the findings. I thought sound evidential falsification of the random wrt fitness qualifier of developmental and genetic variation in evolution was interesting. You say it's just the same old same old same.
Typical. Have a good week.
Comment by Joy — January 20, 2008 @ 10:26 pm
January 20th, 2008 at 11:40 pm
My condolences Doug. I did not think the Giants would do it.
Comment by Bradford — January 20, 2008 @ 11:40 pm
January 20th, 2008 at 11:45 pm
Yes, that was their conclusion.
It's rather hard to parse that sentence, but they did not falsify random mutation with respect to fitness. They eliminated drift as the primary engine of evolutionary change with regards to the traits they studied, and provided a methodology others can use to extend their findings.
Comment by Zachriel — January 20, 2008 @ 11:45 pm
January 21st, 2008 at 8:43 am
There is a Panda's Thumb post about the poor reporting of this paper
http://pandasthumb.org/archive...
Comment by The Pixie — January 21, 2008 @ 8:43 am
January 21st, 2008 at 8:57 am
The news report is only good for alerting people that a new paper is available, and may not always accurately reflect its content or relevance.
Comment by Zachriel — January 21, 2008 @ 8:57 am
January 21st, 2008 at 9:50 am
Thanks for that link, Pixie. Looks as though for this research the press got completely stewed. I have to wonder why that is, given the number of researchers and countries involved. Was there disagreement among them about what the results demonstrated? It certainly seems so.
That said, I'm darned glad RBH expressed the same exasperation I did - we have NOT been waiting 150 years to settle the question of whether evolution is entirely random! And the price of that bridge in Brooklyn just went down 10%. Any takers?
Comment by Joy — January 21, 2008 @ 9:50 am
January 21st, 2008 at 3:20 pm
Interestingly enough, ScienceDaily has removed the most recent (Jan. 20) story - the link now goes to the main page where the story is conspicuously absent. I know that's very likely NOT a result of my notice of the confusions sown here in this thread, though it might be a product of Panda's Thumb notice of those same confusions. Thus far there are 21 comments to the thread, and no one has mentioned contacting ScienceDaily about the problems. No doubt someone did, or maybe the American Technion Society (whose confusing press release ScienceDaily copied verbatim) was notified and withdrew it.
The latest comment at PT (as of now, that's Olorin here) has a broader complaint. The commenter admits s/he is not a biologist, then says s/he went to the actual paper to see what in the world it was claiming. A good idea if you're not scared by reading papers. His/her impression?
Uh, oh. Citing Behe's The Edge of Evolution, the commenter is concerned that this paper might lend credence to the notion that beyond a certain point mutation/ variation has to be guided in order to produce different kinds of organisms. I haven't read Behe's book, so I don't know how accurate that characterization is, though I can certainly see why the confusion sown by the news reports in question might lead to such a concern.
I'm not agreeing - it doesn't appear that this research actually demonstrated guided mutation/ variation, but rather guided development.
The best I could figure from the paper - also not being a biologist - is that variation (even if it's "astounding" in its evolutionary breadth) is irrelevant to the particular organ structure best suited to survival and reproduction in the class of nematodes examined. IOW, no matter how much variation on the theme comes along, they all end up in the same place - and that it's the end result of development that takes it there regardless of how it starts out or how "astonishing" the evolutionary changes in the genes and/or developmental programming might be in different species.
Can I ask critics how, specifically, this multinational research effort could possibly be parsed in support of those non-astonishing incremental changes Zach keeps referring to over evolutionary time frames as how different kinds of organisms evolve? Not just the confusion of bad reporting, but the confusion inherent in the findings themselves - they cited selection, but if there's selection determining the outcome it looks to work backwards in time (given that this does in fact appear to be a top-down imposition, not a bottom-up diversification). At least in regards to roundworm vulvas.
Extrapolating that to giraffe necks would have all giraffe necks over time (per this development) getting longer regardless of what mutations occur or what developmental diversity is engendered - because "selection" somehow decided long necks were the goal. Given, of course, that "selection" must be blamed because it's the only fully deterministic mechanism in the theoretical framework.
Any ideas on how to interpret in light of these findings?
Comment by Joy — January 21, 2008 @ 3:20 pm
January 21st, 2008 at 6:25 pm
I wouldn't worry about that too much. After all, the phenotypic trait in this study that's supposedly most important for fitness - the structure of the vulva - didn't change. Far be it from me to judge the Designer, but her ways would sure be mysterious to me.
Given the directionality of the evolutionary trends, not much more can be said than that some directional force was the cause (I know, that doesn't sound earth-shattering). Which still is an interesting observation, because it might have been random. Maybe selection was involved, maybe not. Maybe biased mutations, maybe not. Upthread I made some more suggestions. To their credit, the authors of the original paper suggest possible future experiments to find out more about the underlying causes. Solid science IMO. It's the ScienceDaily folks who did a less than stellar job. Maybe they're dishonest political hacks, or maybe they're just incompetent.
Comment by Raevmo — January 21, 2008 @ 6:25 pm
January 21st, 2008 at 7:15 pm
Raevmo:
I think (from the little I can parse) the directionality ruled out randomness more than it confirmed selection or suggested directed mutation. The conclusion was that it's not random, that much is entirely clear. It's the cause of non-randomness that's unsettled.
I agree. Want to see this expanded upon, repeated (maybe with some other traits in some other kind of critters). It's an exciting finding that fairly demands further research, a.s.a.p. Who wouldn't want to know?
Neither. The ScienceDaily story was basically a verbatim reprint of the American Technion Society press release dated January 16. In which Professor Benjamin Podbilewicz said,
Simply establishing that "evolution is not random" is a very big development in biological science. Now that this is established, the hows, whys and wherefores are begging to be known. Just think of how much we could do with such knowledge!
I'm glad they pulled the confusing report. I think they realized the importance of the finding, but were unsure of how to parse it in terms of RM-NS (and leery of the ongoing debate). I'd very much like to see what the research teams and their leaders think on that score, even if they all don't agree. It would give a good idea of what follow-up research should look like.
Comment by Joy — January 21, 2008 @ 7:15 pm
January 21st, 2008 at 7:37 pm
Joy:
Exactly.
Damn, you're right. ScienceDaily, if you're reading this, I am very sorry. Don't forget to mention my hugely important work in the future please, pretty please.
You ask me, the cure for cancer is just around the corner.
Comment by Raevmo — January 21, 2008 @ 7:37 pm
January 21st, 2008 at 9:07 pm
Forget mutations for a moment and think about variation.
Giraffe's have an average neck length which varies as a normal distribution around that average. If selection is at work, then those with longer necks will be more successful in leaving offspring. Over many generations, new variations will appear, but now they will be grouped in a normal curve around the longer average neck length. So even though mutations are random, there are enough of them over many generations, that there will simply be normal variation in the population.
What we're attempting to observe is the center of that normal distribution. Will it fluctuate stochastically over time, or will it tend in one direction. Direction implies selection (or constraint).
Darwin proposed his theory of Natural Selection in 1858, and posited that evolution was primarily adaptive, i.e. non-random. The study is interesting, but certainly the headline, New Findings Confirm Darwin's Theory: Evolution Not Random, was overblown. Most everything confirms Darwin's Theory of Evolution.
Google remembers.
Comment by Zachriel — January 21, 2008 @ 9:07 pm
January 21st, 2008 at 9:44 pm
Zach:
So why did they claim it was selection? Why do you put "constraint" in parentheses? That doesn't make much sense in the context of the research itself. Looks to me like there are constraints in action, and selection may play a subordinate role as usual. We don't have any idea right now what those constraints are, several primary mechanisms of selection have been ruled out. Think constraints. What might they be?
And thanks for the cached link.
Comment by Joy — January 21, 2008 @ 9:44 pm
January 21st, 2008 at 10:48 pm
They didn't. They found the change was deterministic, but did not accurately distinguish between selection and constraint (meaning something that reduces the amount of variation on which selection can act).
Trends, Stasis, and Drift in the Evolution of Nematode Vulva Development: We propose that developmental evolution is primarily governed by selection and/or selection-independent constraints, not stochastic processes such as drift in unconstrained phenotypic space.
But as developmental drift indicates there are many ways to produce the same structure, this suggests that generative and developmental constraints are not predominant.
Selection is not ruled out (but likely). They propose further research for making the distinction.
Trends, Stasis, and Drift in the Evolution of Nematode Vulva Development: This approach does not allow a similarly stringent distinction between bias by selection and by selection-independent causes. However, it paves the way for follow-up investigations using other methods, like genetic and population genetic approaches.
Comment by Zachriel — January 21, 2008 @ 10:48 pm
January 22nd, 2008 at 12:24 am
So. Zach is rooting for selection, I'm rooting for directed inheritance. Who is surprised?
Tell me, Zach. Should there be any bias in funding for further researches on either possibility here?
Comment by Joy — January 22, 2008 @ 12:24 am
January 22nd, 2008 at 8:31 am
I'm not rooting for any side, Joy. I'm merely stating what the study's authors thought was more likely, though they were careful not to claim more than what the evidence allows.
The proposed research should distinguish between selection and constraint.
Comment by Zachriel — January 22, 2008 @ 8:31 am
January 22nd, 2008 at 9:54 am
I know, I didn't feel good about that match up the whole game. Eli was just too on. And the Packers were doing stupid things like a 2 yard screen play when they needed 15 on 3rd.
Comment by Doug — January 22, 2008 @ 9:54 am