24 May, 2009

Sunday Sensational Science

Overselling Ida: A Cautionary Tale

If you haven't heard of Ida, the perfectly preserved Darwinius masillae, you've been living in a box. Plenty of information about this remarkable but not utterly revolutionary fossil shall follow. But first, I want to share an illustrative personal anecdote.

A couple of day ago, my coworker, whom I shall call C, asked me if I'd seen the doodle on Google. Indeed I had. And I'd had a brief moment of the warm fuzzies, because it was nice to see that cute little sketch of Ida there in place of the usual logo. Those warm fuzzies turned to the cold ohforfucksakes when C started babbling about her being a "missing link."

This is a normally intelligent man. I rolled my eyes. "No, she's not."

"Yes, she is!"

I was in the midst of Brian Switek's wonderful post on the hype, which will be highlighted below. I'd read several ScienceBlogs posts by then, dissecting the discovery, and I attempted to explain to C that while Ida was awesome, she wasn't the missing link. Indeed, "missing link" is complete bullshit. His response was to inform me that he'd take the word of scientists published in a peer-reviewed journal over what bloggers said any day, and oh yes there are missing links! It didn't matter to him that the peer-reviewed paper had major, major problems, that the whole process had been tainted by publicity stunts, or that the bloggers in question were scientists who know the peer-review process intimately. To him, that process is infallible. Therefore, Ida is the missing link.

So this edition of Sunday Sensational Science is dedicated to C, who reminds me that scientists must resist overhyping their finds in order to score History Channel documentaries, not all peer-reviewed science journals are created equal (although the public doesn't know that), and that having a stable of scientists manning the blogs is a sovereign remedy against sensationalism. Now if we could just get the general public to realize that...

Let's begin with missing links. As in, there are no "missing links":
Again, the press are talking about "the missing link". Let's get one thing clear. There is no missing link. Rather, there are an indefinite number of missing branches. To have a missing link, you need to visualise evolution as a chain. If there's a gap in the chain, then you have a missing link. But evolution, at least at the scale of animals and plants, is mostly a tree.
Keep that in mind as the History Channel, other press outlets, and hysterical creationists endlessly repeat the "missing link" crap. THERE IS NO MISSING LINK. Tattoo it on your hand if you have to.

Now. On to Ida.

Ida has been presented as a near-miraculous superfossil. Our first clue that there was something rotten in the state of Denmark probably should've been the email Brian Switek received:

Late last week I received a rather curious e-mail. It read;


Ground-Breaking Global Announcement

What: An international press conference to unveil a major historic scientific find. After two years of research a team of world-renowned scientists will announce their findings, which address a long-standing scientific puzzle.

The find is lauded as the most significant scientific discovery of recent times. History brings this momentous find to America and will follow with the premiere of a major television special on Monday, May 25 at 9 pm ET/PT chronicling the discovery and investigation.

Who: Mayor Michael Bloomberg; International team of scientists who researched the find; Abbe Raven, President and CEO, A&E Television Networks; Nancy Dubuc, Executive Vice President and General Manager, History; Ellen Futter, President, American Museum of Natural History

"The most significant scientific discovery of recent times", eh? What could it be? Life on Mars? Time-travel? Teleportation? The Higgs Boson? A diet cola that doesn't taste absolutely awful? Well, no. It's all about a little primate from Germany.

When a paper is released in conjunction with a documentary, everybody should put their skeptic's hats on and brace for the worst.

Brian engaged in a bit of prediction:

Consider, for example, the grand claims made about finds like Darwinius. It is being heavily promoted but scientists have not yet had a chance to see the fossil or read the paper describing it. When they get a call from a journalist or are asked their opinion on it, then, it can be difficult to discuss the find because they do not know the details. This can be harmful as it can not only lead to the spread of overblown assertions but it can also make us look foolish if these finds do not turn out to be all they were cracked up to be. This could especially be the case with Darwinius. Though heralded in documentaries and in the news as one of our direct ancestors, it is probably a very interesting lemur-like primate on a different evolutionary branch. I can only imagine the field day creationists are going to have if this is the case, and I am frustrated by the way mass media outlets manage to bungle genuinely interesting scientific discoveries.

He was so right.

The paper was published to a media frenzy. The drama got so bad that some science bloggers were forced to resort to emergency satire:

Yesterday, the entire world changed noticeably as the media, accompanied by some scientists, unveiled a stunning fossilised primate. The creature has been named Darwinius masillae, but also goes by Ida, the Link, the Chosen One and She Who Will Save Us All.

The new fossil is remarkably complete and well-preserved, although the media glossed over these facts in favour of the creature's ability to cure swine flu. Ida was hailed as a "missing link" in human evolution, beautifully illustrating our transition from leaping about in trees to rampant mass-media sensationalism.

No one's disputing that Ida's a remarkable find. PZ sang her praises thusly:

What's so cool about it?

Age. It's 47 million years old. That's interestingly old…it puts us deep into the primate family tree.

Preservation. This is an awesome fossil: it's almost perfectly complete, with all the bones in place, preserved in its death posture. There is a halo of darkly stained material around it; this is a remnant of the flesh and fur that rotted in place, and allows us to see a rough outline of the body and make estimates of muscle size. Furthermore, the guts and stomach contents are preserved. Ida's last meal was fruit and leaves, in case you wanted to know.

Life stage. Ida is a young juvenile, estimate to be right on the transition from requiring parental care to independent living. That means she has a mix of baby teeth and adult teeth — she's a two-fer, giving us information about both.

Finds like Ida are extremely rare, and she's justly being celebrated as an important find. But the overselling is, ironically, selling her short. It's like promising someone a Ferrari and delivering them an Altima. That's where buyer beware comes in handy, and Brian Switek's done an excellent job kicking tires on this one:

Some researchers have long maintained that adapids are better candidates for the ancestors of anthropoids, with Philip Gingerich (one of the authors of the Darwinius paper) being a vocal proponent of this view. It is not terribly surprising, then, that the authors of the Darwinius paper posit that adapids are more closely related to anthropoids than tarsiers and omomyids, and they rely on two tactics to make their case.

The authors of the paper try to frame their hypothesis in a historical manner. They claim that adapids have been barred from a close anthropoid relationship on the basis of soft-tissue characteristics that do not fossilize. This would mean that the association between omomyids, tarsiers, and anthropoids would hang by a nose, but this is not true. As reviewed in popular books like Chris Beard's The Hunt for the Dawn Monkey and technical volumes like Anthropoid Origins, the relationship between omomyids, tarsiers, and anthropoids is based upon a wide array of fossil and neontological data. I can't imagine why the authors of the new paper would suggest otherwise unless they were trying to construct a false historiography in order to show their fossil in a better light.

This shoddy scholarship is matched by a weak attempt to show that Darwinius has more anthropoid-like traits than tarsiers or omomyids do. In order for the authors of the paper to make a convincing case they would have to undertake a careful, systematic analysis of the anatomy of Darwinius in comparison to other primates, yet they did not do this. Instead they combed the literature for 30 traits that might help ascertain the placement of Darwinius in the primate family tree and filled in whether each trait was present or absent in Ida's skeleton.

That's the post I was reading when C started spouting off about peer-review vs. bloggers. I sent him the link. Really, when you think about it, science bloggers engaging in research blogging are peer-reviewers. And then you have Carl Zimmer, one of our best science journalists, engaging in a little peer review of his own:

It finally got to the point where I found myself dispatching emails to two prominent primatologists–John Fleagle of SUNY Stony Brook and Chris Beard of the Carnegie Museum–to see what they thought.

Both researchers agreed that it was a lovely fossil, in terms of its exquisite preservation. “It’s really wonderful,” Fleagle said. It’s got bones, fur, and even its last meal in its stomach. Fleagle observed that it will be possible to learn many details about the biology of early primates from Darwinius, down to the stages by which it teeth erupted.


...I asked what Fleagle and Beard thought about the actual argument in the paper, which has to do with where humans, apes, and monkeys (known as anthropoids) fit in the primate family tree. Some of the co-authors on the new paper have argued in the past that an extinct group of primates called adapiforms gave rise to anthropoids. Others have favored a common ancestry with small primates known as tarsiers. (Laelaps has a nice history of the debate.) The authors of the new paper argue that Darwinius is an adapiform, but it also has traits that link it with anthropoids. So, according to them, it’s an early relative of our own anthropoid lineage.

Both Fleagle and Beard were not impressed with this argument. Fleagle observed that, ironically, most of the evidence presented in the paper is old news. Except for the ankle and a few other traits, most of the traits offered to link adapiforms to anthropoids “have been known for decades,” said Fleagle. It’s nice to have those traits all in one primate fossil, but they don’t advance the debate. Fleagle is intrigued by the anthropoid-like ankle of the fossil, but he also notes that it’s “roadkill,” flattened down to a 2-millimeter pancake. He wonders whether their interpretation of the ankle will hold up to scrutiny.

Beard has similar things to say via email.

I’ve been deluged today by journalists regarding this. It is a marketing campaign for the ages. The fossil is nice because it is so complete, but it is a rather vanilla-flavored adapiform that does not differ appreciably from other members of that well-known group of Eocene primates…

Beard was also puzzled that the authors did not compare Darwinius to an important early anthropoid fossil Beard found, known as Eosimias. In fact, he was underwhelmed by the entire comparison of Darwinius to other primates (a phylogenetic analysis):

The phylogenetic analysis is not very complete, and I would certainly interpret many of the characters they do cite very differently than they do. But one of the most shocking things of all about the technical paper is that they found room to cite 89 references, but there is not one mention of Eosimias to be found there. This is bizarre indeed. In a paper that purports to tell us something about anthropoid origins, the authors have conveniently ignored the single most significant fossil that has been published to date. Incomprehensible.

From all available evidence, it seems the authors of the paper were more interested in trimming facts to fit their theories than in good science, and a lot of that motivation probably came from their chance at fame and fortune. It's a shame. Carl's right: science is being held hostage:
So, to recap: it appears that both PLOS and Atlantic Productions did not give journalists any time to consult with outside experts before launching a major press conference with a huge blitz of media attention. In other words, science writers who were trying to do their job well and responsibly were actively hindered. Those who declared ridiculous things, such as claiming that human origins were now solved once and for all, were not.
This, my friends, is not the way to do science. PZ points out some of the many issues:
This is not helping. It is inflating a good discovery beyond all reason, and when the public hears the creationists declare that it's one fossil of a monkey-like creature, and they're right, it's going to damage the credibility of science.

Seed Media has a bit of a scoop: they've got an interview with a PLoS One editor, a History Channel executive, and Jørn Hurum, the scientist behind all the promotion. It's appalling. They're bragging about how a "production company got in on the ground floor". Shall we anticipate the brave new world when paleontologists have to beg for McDonald's happy meal tie-ins to get funding?

Ida deserved better than this. She's an amazing little creature, and she's getting lost in hype. Thankfully, Brian Switek's rescuing her from the frenzy, and helping her demonstrate what she has to teach us:

First, how do we know that Ida was a female? It all comes down to a missing penis bone, or baculum. Many, if not most, mammals have a penis bone, and in fact our species is one of the "oddballs" in that males of our species do not. Take a look at the restoration of the transitional pinniped, Puijila, that was announced a few weeks ago. See that long bone sticking out from in front of its pelvis? That's a baculum, and the presence of such a bone indicates that this specimen of Puijila was a male.

While our species might not have a baculum, other primates do, including fossil ones. Darwinius lived alongside another kind of lemur-like adapid primate called Europolemur, and fossils from the same Messel shales show that male Europolemur had large baculums. Given their close evolutionary relationship between Europolemur and Darwinius it can be reasonably assumed that male Darwinius had baculums, too, but Ida's skeleton does not have a penis bone. Is it possible that this specimen of Darwinius could have been, pardon the expression, dis-membered sometime after death and before fossilization? It is possible, but given the exceptional preservation of the fossil, including gut contents and a body outline created by bacteria, it is doubtful. The lack of a baculum attached to this fossil makes it highly probable that Ida was indeed a female.


Even though I have been critical of the way this entire "primate roll-out" has been handled, I have tried to stress how amazing a fossil Darwinius is. The sex and age of a fossil might seem like unimportant matters, but how often do we get such a clear window into the biology of an extinct species? Right now the public is still being deluged with the message that Ida is the "missing link", but I hope that what Ida's skeleton can actually tell us about how she lived and died receive greater attention as we continue to discuss her bones.

Scientists like Brian will ensure that Ida doesn't get lost in the hubub. And maybe, just maybe, this is a precious teachable moment. Wouldn't it be lovely Ida not only taught us about the evolution of primates, but helped the general public understand good science versus bad or biased science? She might even save us from PZ's nightmare of McDonalds-sponsered paleontology.

That would make her a miracle indeed.


Efrique said...

One of the problems of the "missing link" idea is it makes people keep thinking of fossils as having to fit into some "sequence" - this, then this, then this**, when in fact if you're thinking about a bush with lots of branches, there's no sequence, there's no "order". These two are similar, that one dates earlier, they share a common ancestor probably back there somewhere, but it is likely that neither is a direct ancestor of any other fossil we have - even though both may be very similar to direct ancestors of fossils we have.

For example, Tiktaalik is often painted in this way - even as an ancestor for all tetrapods - when it almost certainly isn't. At best it's likely a close relative of whatever such a common ancestor would be (a short side-branch off), and proof that there were indeed creatures that we would classify as fish which bore many many characteristics of (and strong resemblances to) amphibians, suggesting that the common ancestor to all tetrapods that would have looked something very like Tiktaalik.

I can understand a newspaper wanting to avoid some of the necessary locution, but it's unforgiveable in a scientist publicizing their work.

** and if you're thinking "sequence" the next thing is you're thinking "what's at the end?/what's the goal?/ what's the purpose?". BAD.

marie said...


We have just added your latest post "Sunday Sensational Science" to our Directory of Science . You can check the inclusion of the post here . We are delighted to invite you to submit all your future posts to the directory and get a huge base of visitors to your website.

Warm Regards

Scienz.info Team