Disputes and controversies in science are always a good thing. They're fun to read about (and to write about), and they're bellwethers of the health of the enterprise. Moreover, they tend to stimulate thought and experimentation. Whether scientists are bickering about evo-devo, or about stem cells in cancer, or about prebiotic chemistry, and whether or not the climate is genial or hostile, the result is valuable.
Now of course, some controversies are invented by demagogues for political purposes. The dispute in such cases is far less interesting and clearly less profitable, even if participation by scientists is necessary.
This week, two papers in Nature weighed in on a major scientific controversy that has its roots in pre-Darwin embryology, fueled by some gigantic scientific personalities and even tinged with what some would call fraud. This intense scientific dispute spawned a sort of doppelganger, a manufactured controversy that is just one more invention of anti-evolution propagandists. The Nature cover story gives us a great opportunity to look into the controversies, real and imagined, and to learn a lot about evolution and development and the things we're still trying to understand about both.
The scientific dispute is an old one, dating to when scientists first began to study embryonic development in earnest. Embryologists like the great Karl Ernst von Baer noticed that the embryos of very different animals often looked so similar that they could hardly be distinguished from each other. A chicken embryo, at some point, looks an awful lot like a human embryo. What does this mean? Two schools of thought (roughly speaking) entered into competition, with evolution as the major subtext. One set of ideas envisioned development as recapitulation: development was a sort of re-play of evolution, with the organism recapitulating its evolutionary history as it took shape. Recapitulation theory was the brainchild of Ernst Haeckel, whose view of development was codified as his Biogenetic Law and sloganeered as "Ontogeny recapitulates phylogeny." Against recapitulation were the views of von Baer and others; von Baer formulated his own set of laws, the third of which repudiates recapitulation rather directly. Everyone agreed that embryos of different animals often looked quite alike; the dispute was about what this meant. And it seems that those who opposed evolutionary explanations (like von Baer) were eager to point to difference and divergence during development, while those who championed evolutionary views wanted to emphasize the shocking similarities between, say, chickens and mammals when compared at key embryological junctures.
Haeckel, famously, went on to point to those similarities as evidence for common ancestry and, infamously, to create a certain illustration of that evidence. His picture, thought by even some embryologists to be partly fraudulent (more accurately, "doctored"), is now a staple of anti-evolution propaganda. You can read all about that elsewhere; suffice it to say that Haeckel's drawings have long since been "corrected" without creating any problems for evolutionary theory. (For a much more detailed treatment of this saga, see Richardson and Keuck, Biological Reviews, 2002.)
But interestingly, the debates about recapitulation morphed (wink) over the years into a distinct but related disagreement about whether animal development passes through a stage that is common to - or typical of - the lineage of the organism. Because although Haeckel's recapitulation idea didn't survive, it remained clear that development seemed to reflect evolutionary commonalities. Consider the photos on the right. (The figure was created by Michael K. Richardson, who led the research group that critiqued Haeckel's drawings in 1998.) While the various embryos shown all end up looking quite different - looking like the adult form, in other words - they seem to "start" at a place that's notably similar. (Compare the embryos in the first row.) That starting point is not the beginning of development, and in fact those different kinds of embryos got to that starting point via rather different beginnings. In other words, it seems that animal embryos pass through roughly three phases of development: an early phase that can vary from group to group (say, between birds and mammals), a late phase in which group-specific forms are established, and a middle phase that is eerily similar among groups. That middle phase has come to be known as the "phylotypic stage" of development, meaning that it is a stage at which the embryo looks like a typical example of its evolutionary group. For insects, this is thought to be the "extended germband" stage; for vertebrates, it's roughly the tailbud stage. The point is that there is a middle phase of development during which animal embryos of varying morphological destinies look very similar, even if their earlier stages seemed very distinct. This model of developmental trajectories, compared across groups, is known as the "hourglass model," nicely depicted by Richardson and colleagues in the cartoon on the left.
Why all the controversy? Well, the disputes all seem to be related to the fact that the model is mostly descriptive. And so, one criticism is that the model is based on what embryos look like, and not strongly anchored in carefully-defined and -measured characters. Moreover, some critics have noted that the comparisons were often restricted to popular laboratory species, such that when the analysis was expanded to include a broader set of species, the similarities in the waist of the hourglass become less striking. In other words, the dispute centered on the basis of the model. Critics were disputing the very existence of the phylotypic stage.
Oh, and while this interesting scientific debate was ongoing, some propagandists were shadowboxing with Haeckel's ghost, shrieking about fraud while creating in the minds of their dupes the illusion of a different debate: one about whether development and evolution are conceptually linked. Along the way, these busy demagogues suggested that the phylotypic stage is an illusion, cherry-picking their data more shamelessly than Haeckel ever did. In any case, these folks were exploiting the real scientific dispute: whether the phylotypic stage can be defined more rigorously, in a way that links the similarities (whatever they are) to common ancestry.
And that brings us to the cover story in this week's issue of Nature. The cover image depicts a version of Haeckel's infamous illustration. The issue includes two reports, very different in their approach and in the animals they examined. Both reports provide striking support for the hourglass model, by showing that the phylotypic stage is indeed characterized by distinctive and fascinating patterns of gene expression. Part II will explore those two papers.
---
Image credits: 1) embryo images from Gilbert, Developmental Biology, 6th Edition, online at PubMed; 2) cartoon from Richardson et al. 1998.
46 Comments
Matt G · 10 December 2010
Why are there no Wikipedia entries for phylotypic stage and hourglass model? And why can't Nature let a few miserable articles go for free?
Troy Britain · 10 December 2010
Ichthyic · 10 December 2010
because wiki is not the be all and end all of human knowledge?
this is a close as you're gonna get to coverage of this stuff on wiki:
http://en.wikipedia.org/wiki/Evolutionary_developmental_biology
Ichthyic · 10 December 2010
oh, and while lambasting of overpriced journals is a great thing, I would also recommend putting some effort into supporting the open journal movement:
http://www.doaj.org/
http://www.openarchives.org/
Ichthyic · 10 December 2010
...and here's a previous, FREE paper on the subject that goes into a bit more of the basics:
http://rspb.royalsocietypublishing.org/content/270/1513/341.full.pdf
which is an example of what Steve refers to with: "In other words, the dispute centered on the basis of the model. Critics were disputing the very existence of the phylotypic stage."
Ichthyic · 10 December 2010
...one last item, Jerry Coyne also discusses this work in case more coverage is desired:
http://whyevolutionistrue.wordpress.com/2010/12/08/of-cats-and-men-old-genes-give-embryos-an-hourglass-figure/
Ichthyic · 10 December 2010
*grr*
thoughts pop in one at a time in my head these days.
...ever wonder why PZ's blog is named what it is?
http://en.wikipedia.org/wiki/Pharyngula
which, btw, is probably also why you can't find a specific "phylotropy" wiki page.
SteveF · 10 December 2010
Ichthyic · 10 December 2010
*doh*
right.
though Jerry does weigh in directly on the subject in that thread.
Matt G · 10 December 2010
Ichthyic · 10 December 2010
but have never come across the hourglass model or phylotypic stage.
It's probably because the terms are kind of cross-disciplinary, as it were.
The hourglass model is much older than the modern evo-devo synthesis; the term is not usually referred to outside standard developmental biology texts, so you won't run into it outside of discussions of exactly what these papers are about, or basic developmental biology courses.
"phylotypic" is not actually a recognized stage of development, but is more of a descriptive thesis from a more purely evolutionary discipline that could apply to several stages of development (though most often and historically applied to the pharyngula stage), so you won't run into it commonly while perusing literature on developmental biology.
so, without a basic college level course in either evolutionary biology or developmental biology, I could easily imagine not running across the terms; I hadn't even thought about it.
You mean there’s more?!
why, yes! shocked me too.
;)
TomS · 11 December 2010
Does this phylotypic stage - the narrowest point of the "hourglass" - represent, more or less, something like the phylum (or thereabouts) throughout the animal kingdom? I'm wondering whether one could tag a certain level of taxonomy as being something like:
Organisms which share a common phylotypic stage belong to the same phylum (or whatever).
fnxtr · 11 December 2010
What metrics would you use, TomS? Seems to me there'd be a lot of differences even in "common" phylotypic stages.
Jim Thomerson · 11 December 2010
P. Z. Myer's illustrated pharyngula stage is clearly identifiable as a minnow, family Cyprinidae (zebra danios are minnows). So it is considerably more specific than phylum. I'm not a minnow expert, so I cannot say if it is identifiable as a member of some sub group of minnows.
David Fickett-Wilbar · 11 December 2010
ravilyn.sanders · 12 December 2010
Re, Hourglass graphic: I am not a biologist. I don't get the hourglass graphic. I would have thought the zygotes would be on top, and after the constriction (the phylotypic stage) we would have different species at the bottom. Hey? Who flipped my hour glass?
Re, Cost of journals: The cost of transmitting, replicating and sharing information has dropped several orders of magnitude over the last few decades. But the cost of collecting and compiling information has stayed constant or has gone up. We still don't have a good model for paying for the info collection. If the consumers of info could somehow come up with a model to pay for the information creation a priori, may be we would be able to take full advantage of the ever decreasing cost of distribution.
At least in the academics most people who create the information are the graduate students and their tenured professors. They might actually think the info has been "paid for" and be willing to share it freely. But publishing it in a blog does not have the same prestige of publishing in "Nature". Why is that?
Because we are acting like female sage grouse birds who choose to mate with the male most admired by most other female sage grouse birds in a lek!
If we collectively come up to fund a web site to publish such articles, and cite papers in that site in our papers, we could print our own money. The coin of our realm is not a federal currency note that would buy you a loaf of bread in the grocery store. It is prestige, it is recognition. We create it. We, collectively, are our own federal reserve. The cost of distribution has fallen so much we can finally dispense our recognition and prestige directly without mediating it through some currency.
I think the readers of this group would be the one among all other sciences to get it.
Steve Matheson · 12 December 2010
Steve Matheson · 12 December 2010
I share everyone's frustration with Nature and its opposite-of-open-access stance, and you may have heard of large boycotts of Nature journals for related reasons. Most maddening is their refusal to provide open access to the archive after any amount of time; all the Cell journals do this, for example. To be fair to Nature, they do have some nice collections (called Specials) with free stuff, organized by topic, and their Scitable site is exemplary. (Scitable includes a lot of free articles from Nature journals.)
As for ravilyn's suggestion: I think that's already being done. Check out PLoS and BMC collections for prominent examples.
Shebardigan · 12 December 2010
Ichthyic · 12 December 2010
Scitable includes a lot of free articles from Nature journals.)
most aren't though, from my recollection.
it's very hit and miss.
I'd just remind folks to support open access, and the relevant links I posted are earlier in this thread.
one of these days, I'm some pirate somewhere is going to get interested enough to blow the lid off journal privacy, and then the publishers of these journals are REALLY going to regret not thinking to reduce the costs to consumers.
I'm actually quite surprised it hasn't already happened.
darwinism.dogbarf() · 13 December 2010
The hourglass shape is obviously a product of intelligent design and not Darwinism. How could all embryos converge to this particular shape in a cognitively isolated environment. Haeckel's biogenic law would appear provide a Darwinian explanation but even PZ Myers admits it fails. There is no way embryonic developent could work this way by pure random chance without cognitive input.
Joe Felsenstein · 13 December 2010
Joe Felsenstein · 13 December 2010
Oops, typing mistake: “But the earliest stages ...”
Joe Felsenstein · 13 December 2010
Oops-squared: “But the very earliest stages ...”
harold · 13 December 2010
I'm eagerly awaiting part 2.
We all already realize that common developmental features are some of the strongest evidence for common descent, e.g. Hox genes.
Whether a definable stage of vertebrate development is "even more conserved" on grounds other than morphology seems to be the point that this article is building up to, and that seems highly plausible for the reasons Joe Felsenstein points out, but since the non-morphology stuff isn't up yet, it's hard to say much.
eric · 13 December 2010
Wouldn't it be more of an "ant shape" than an hourglass shape? After all, I imagine pretty much every vertebrate zygote is similar. So it starts off similar, gets dissimilar, goes through talibud stage, gets dissimilar again.
darwinism.dogbarf() · 13 December 2010
mrg · 13 December 2010
DS · 13 December 2010
John Kwok · 13 December 2010
Ichthyic · 13 December 2010
You assume that these changes occur by the random forces of natural selection and not the information in their genomes.
utter gibberish.
harold · 13 December 2010
Ichthyic · 13 December 2010
Dogbarf is presumably a parody
for their sake, I actually hope so.
JLT · 13 December 2010
But it is a great point in favour of my hypothesis that IDist are pro-ID because they are inherently unable to grasp the concepts of analogy and metaphor. Just out of interest, do you also believe that because histograms of samples with a normal distribution like body height (example) can be described as bell-shaped that body height is intelligently designed?
darwinism.dogbarf() · 14 December 2010
Ichthyic · 14 December 2010
You actually tried o make an argument. As for you example, I would say since a normal distribution indicates randomness, the heights of Swedes has not been intelligently designed.
how do you know it's not being cleverly hidden in the noise?
I hear the Abrahamic god is quite the trickster like that, getting his minions to hide fossils everywhere, and working so hard to make it look just like natural selection can explain the changes in so many traits, and the evolution of so many species.
*shakes fist*
how are we to know this isn't all a trick, and the universe wasn't created last Thursday?
harold · 14 December 2010
darwinism.dogbarf() · 14 December 2010
Ichthyic · 14 December 2010
Well, we don’t, but that is a question of religion, not science.
then how does religion answer it?
The science of intelligent design can only infer
not much of a "science" then, eh?
Have you tried asking him?
who? Ganesh?
pick one:
http://www.magictails.com/creationlinks.html
harold · 14 December 2010
phantomreader42 · 14 December 2010
Flint · 14 December 2010
phantomreader42 · 14 December 2010
Flint · 14 December 2010
darwinism.dogbarf() · 15 December 2010
Stanton · 15 December 2010