November 6, 2008
Bone Wars in the Blogosphere
When a scientific paper is published, it is not the last word on the topic. It is truly only the beginning, and that new research becomes widely available for debate and discussion. Normally comments are traded between experts, and arguments take place in the halls of symposia, but blogs and open access publishing allow the public a unique look into how scientists react and respond to published research.
Two weeks ago I wrote about a newly-announced predatory dinosaur named Aerosteon, described in the open-access journal PLoS. For years, the scuttlebutt among paleontologists was that these fossils were a significant find, which the PLoS paper confirmed, but not everyone was entirely bowled over by the paper.
At the blog SV-POW!, which specializes on the weird vertebrae of sauropods, paleontologist Matt Wedel wrote a detailed critique of the Aerosteon paper. Beyond debating the anatomical analysis of the new dinosaur, Wedel charged that those who described Aerosteon had not properly cited (and even unfairly criticized) previous studies on air sacs in dinosaurs.
Particularly perplexing was a quote from Paul Sereno, one of the scientists who wrote the Aerosteon paper, who said “The fossil [Aerosteon] provides the first evidence of dinosaur air sacs, which pump air into the lungs and are used by modern-day birds.” [emphasis added]
As Wedel aptly pointed out in his first post on the subject, air sacs in dinosaur bones have been recognized for over 100 years, and in the past decade a more detailed research program have centered on these features (on which Wedel himself is an expert). Even as recently as 2005, a well-publicized paper was published about air sacs in the predatory dinosaur Majungasaurus (then called Majungatholus). Aerosteon is cool, but it’s not the first time these features have been seen by a long shot.
Paul Sereno and his co-authors have not taken Wedel’s criticisms lightly. In a response posted to the discussion boards at PLoS, Sereno wrote:
In two trackbacks to our paper, Matt Wedel offers a misleading, longwinded, ad hominen critique of this paper on the new theropod dinosaur, Aerosteon riocoloradensis, and the significance of its pneumatic features. Some personalized aspects of the commentary and erroneous claims push the limits of the “good practice” guidelines posted for commentary in this journal.
Sereno goes on to describe what he and his fellows authors aimed to do in the paper, but ultimately he remarks that he does not feel “personalized, ad hominem blogs like Wedel’s advance scientific understanding or enhance collegiality.”
This response troubles me for several reasons.
First, Wedel’s response appeared on his own blog and is not subject to whatever “good practice” guidelines might apply to discussion on the PLoS website. Even so, the response of Sereno deems Wedel’s critique an “ad hominem” attack (meaning directed against the authors rather than their research) without illustrating why this is so.
In fact, none of the specific issues Wedel brought up in his post were addressed in Sereno’s reply. Sereno’s reply, for instance, states that the authors of the Aerosteon paper strove to “Cite the literature thoroughly and fairly (95 citations).”
Ninety five citations is an imposing number, but sheer volume of references alone does not constitute an argument. It is what is said about those citations that matters, and none of the points Wedel brought up were responded to.
Wedel, in turn, has posted a reply to the latest dispatch. As he freely admits, he did speculate on why, in his view, some of the work on air sacs in dinosaurs had been misinterpreted in the Aerosteon paper. This is likely what the authors of the paper considered to be an ad hominem attack, even though no specific response to Wedel’s speculations were provided. After a criticism of the terse response, Wedel concluded:
If someone brings a fact-based critique against your work, rebut them with facts or not at all. Calling names just makes you look weak and gives the impression that you have no factual case to pursue. My critique of the Aerosteon paper is “longwinded” only because it is so thoroughly documented. Sereno tries to paint it as a content-free exercise in pique–which is a pretty fair description of his own response. The irony could hardly be any richer.
I lack to expertise to be any type of arbiter on the technical issues, but the fact of the matter is that Wedel wrote a very detailed critique (practically a paper by itself) of the Aerosteon research in the spirit of scientific discourse. The reply from the authors of the paper, by contrast, addressed none of his points and went so far as to try and discredit Wedel’s criticisms by belittling the fact that he shared his arguments with the interested public.
I also beg to differ with the authors of the Aerosteon paper that Wedel’s work (and science blogs in general) do not advance scientific understanding. I learned far more from reading Wedel’s point-for-point discussion than I would have been able to on my own. It makes a technical discussion otherwise held between experts, largely inaccessible to the public, available to everyone who is interested.
Science blogging still has a bit of a “Wild West” atmosphere where ethics and niceties are still being worked out, but it does provide a powerful tool to discuss and respond to new research. This is especially important when there are aspects of new papers that appear to be false or can be debated. This provides the public a view into how we come to understand what we know about the natural world, and I hope that the authors of the Aerosteon paper make the most of the opportunity to let us in on the scientific discussions in a more substantive reply.
Sign up for our free email newsletter and receive the best stories from Smithsonian.com each week.
5 Comments »
RSS feed for comments on this post. TrackBack URI
It is poor journalism to attribute every statement made by reporters in the haste of a news announcement to the authors of a scientific paper; the paper or actual press release should be checked for substantiation. Brian Switek has added quotation marks to a comment made by a reporter (Orozco)—a statement that I never made. Here is what Brian said I said:
Particularly perplexing was a quote from Paul Sereno, one of the scientists who wrote the Aerosteon paper, who said “The fossil [Aerosteon] provides the first evidence of dinosaur air sacs, which pump air into the lungs and are used by modern-day birds.” [emphasis added]
We never claimed that, and I never said that to any reporter. We reviewed all of the fossil evidence for air sacs in the paper and present the first well-documented and unambiguous evidence for clavicular and abdominal air sacs, both of which are ventilatory air sacs in extant birds.
Wedel himself has offered two contradictory interpretations of the origin of pneumatic features along the axial column in nonavian dinosaurs. We regard this evidence as ambiguous as well, given the lack of close morphological correlates to the continuous pneumatic openings in the centrum, the extreme modification of the posterior axial column and pelvis, and the absence of any pneumaic hiatus in extant avians (see paper for further discussion).
Brian Switek dismisses the short comment we posted in response to Wedel (in comments associate with the paper), unaware that this is only a shorter version of a longer response (http://www.paulsereno.org/blog-1.html).
We support vigorous debate in science; we do not regard as productive personalized, ad hominem attacks that in places deviate from the facts, issues and hypotheses under investigation.
[...] Switek weighs in on the back-and-forth in a post on Smithsonian.com’s Dinosaur Tracking [...]
[...] part of an underinformed journalist. At least, it’s not in the official press release, and Sereno denies ever saying it. It is interesting that in the National Geographic story it is identified as something Sereno said [...]
Wedel himself has offered two contradictory interpretations of the origin of pneumatic features along the axial column in nonavian dinosaurs.
Really? I would be very interested to see that statement backed up with quotes from the relevant papers, which are not identified here.
In all of my work on pneumaticity and air sacs in dinosaurs, there have been three big areas of uncertainty:
(1) is it possible for the entire vertebral column to be pneumatized by diveticula of cervical air sacs?
(2) when did abdominal air sacs first evolve in archosaurs, and how many times?
(3) how much weight should we give to vertebral fossae, which can be associated with many kinds of soft tissues (cartilage, muscles, fat, and air sacs) and are not always diagnostic of pneumaticity?
(1) In 2000 my coauthors and I posited that because the posterior vertebral column of birds is pneumatized by diverticula of the abdominal air sacs, the presence of pneumaticity in the posterior dorsals, sacrals, and caudals of sauropods is evidence that they had abdominal air sacs. Although Sereno et al. (2008) did their best to discredit it, the evidence for abdominal air sacs from vertebral pneumaticity is very strong, especially since O’Connor and Claessens (2005) showed that when the posterior part of the vertebral column is pneumatized in birds (it is in some taxa and not in others), it is always by diverticula of abdominal air sacs and never by diverticula of cervical air sacs. That work had not been published when I wrote my 2003 paper, so I proposed pneumatic hiatuses as an additional line of evidence to potentially sort abdominal and cervical sources of pneumatization. The fact that we have not identified any pneumatic hiatuses in non-avian dinosaurs to date is meaningless; they are not present in many birds, and there are several good reasons to expect them to be rare in non-avian dinosaurs, as I explained in the original paper and recently reiterated. This is one of those classic cases in which absence of evidence is not evidence of absence: if we find a pneumatic hiatus in a non-avian dinosaur, it will provide additional evidence for the already robust hypothesis that saurischians had abdominal air sacs; if we don’t find one, that hypothesis is still supported by other lines of evidence.
(2) Inferences about pneumaticity and air sacs are asymmetric. If we find certain patterns of pneumaticity, we can infer that certain air sacs were present, but if we don’t find those patterns, we can’t assume that air sacs were absent. Some birds (loons, penguins, kiwis) have no postcranial pneumaticity despite having the full kit of avian air sacs. This makes it tough for us to determine when the air sac system first evolved. There is good evidence for both cervical and abdominal air sacs in non-avian theropods, sauropods, and pterosaurs, but to date only equivocal evidence in ‘prosauropods’ and none at all in ornithischians. So there are at least three possibilities for the origins of the air sac system:
(A) once at the base of Ornithodira;
(B) twice independently in pterosaurs and saurischians;
(C) three times independently in pterosaurs, sauropods, and theropods.
(A) requires only one origin, but also requires that ornithischians had air sacs but never pneumatized their postcranial skeletons despite having about 160 million years of opportunities. (B) avoids that problem, at the expense of requiring at least two origins of an air sac system. (C) basically means that every time we find pneumaticity in fossil archosaurs, we infer a separate origin of an air sac system. Personally, I think it is no great stretch for ‘prosauropods’ to have had air sacs and just not gotten around to pneumatizing the postcranial skeleton, so I prefer (A) or (B). All three are contradictory hypotheses, but only about the inferences we draw from the available data, and not about the solidity of those data.
(3) Were ‘pneumatic’ fossae actually pneumatic? This is an important question because it bears on the origins of postcranial pneumaticity and air sacs. As we go down the tree toward basal dinosaurs and basal archosaurs, the problem is not that osteological correlates of pneumaticity disappear entirely; it’s that they become so indistinct that they are no longer diagnostic for pneumaticity alone. For example, most prosauropods have shallow indentations on the sides of their vertebrae, but do these fossae represent remodeling of bone by pneumatic diverticula, or were they filled with cartilage or fat like similar fossae in extant crocodilians and mammals? If the fossae are pneumatic, we can be more certain that prosauropods had air sacs, and we can more confidently reject hypothesis (C) above. In my earlier papers I was more bullish on pneumatic fossae than I have been lately, mostly because I’ve seen some very distinct fossae in crocodilian vertebrae that would probably be identified as pneumatic if they turned up in a dinosaur. Still, these things have to be evaluated on a case-by-case basis, and no blanket conclusions are possible. Some of the putatively pneumatic fossae of basal taxa were probably pneumatic, and some almost certainly were not. What we need most is new tools and techniques that would help us differentiate the two sets.
So have I offered two contradictory interpretations? I shared everyone’s uncertainty about cervical pneumatization of the posterior vertebral column–at least until O’Connor and Claessens (2005) sorted that out. I have admittedly waffled a bit on fossae, but only because it’s a hard problem that desperately needs to be answered (and answered correctly). As for the single vs. multiple origins of air sacs, I have discussed the alternatives as openly and forthrightly as I could. I’m not sure to which of those sources of uncertainty Dr. Sereno is referring, and I would appreciate clarification–hopefully fully referenced.
We regard this evidence as ambiguous as well,
Overstatement alert! I’ve never regarded the evidence for postcranial pneumaticity in sauropods and non-avian theropods to be ambiguous. Sauropods and most non-avian theropods had pneumatic vertebrae. There is no other soft-tissue system that produces the distinct cavities and chambers we find in their vertebrae, and no one has seriously challenged the pneumatic interpretation in well over a century. The question is not whether sauropods and non-avian theropods had pneumatic vertebrae or even air sacs–the questions are all about when those systems first kicked in, and how strong the evidence has to be before we feel confident about identifying those origins.
given the lack of close morphological correlates to the continuous pneumatic openings in the centrum,
This is misleading. Many birds do have a continuously pneumatized vertebral column, although the pneumatic foramina may be inside the cervical rib loops where they are hard to see, or up on the neural arches rather than on the centra. And some birds do have foramina on the centra right down the column. This is discussed at greater length here. I also urge everyone to read O’Connor (2006) on the close correspondence of pneumatic features among birds and non-avian saurischians.
the extreme modification of the posterior axial column and pelvis
As Dr. Sereno explains on his blog, “The point we made concerning the abdominal air sac in nonavian theropods is that it could not possibly have occupied the volume it does in extant avians under the posterior axial column; the articulated preservation of pubes in Aerosteon and posterior dorsal and sacral centra nearly close the pelvic outlet.”
On the other hand, Sereno et al. (2008) argue that Aerosteon had abdominal air sacs despite its narrow pelvis, so why not other saurischians? Second, we could wrangle all day about whether we think that saurischians had enough room for abdominal air sacs; what is more important is that most saurischians have patterns of pneumaticity that are diagnostic for abdominal air sacs.
and the absence of any pneumaic hiatus in extant avians (see paper for further discussion).
I’m sure that Dr. Sereno meant to say “the absence of any pneumatic hiatus in non-avian dinosaurs”, since pneumatic hiatuses were first identified in birds and are particularly well-documented in chickens. In any case, the importance of pneumatic hiatuses has been grossly overstated; as I have repeatedly explained, they will be at best an additional line of evidence for bird-like air sacs in non-avian dinosaurs, if we find any. If we don’t find any, they’re nothing. Go look at cranes, pelicans, ostriches, and a host of other birds, and you’ll find no pneumatic hiatuses in most adults. Even in chickens, where pneumatic hiatuses were first identified, they are not present in all individuals. Sereno et al. (2008) may consider the available sample of well-preserved juvenile saurischians to be large enough to feel confident about the absence of pneumatic hiatuses. I don’t, but time will tell.
[...] week I wrote about the controversy over the new PLoS paper describing the dinosaur Aerosteon. In the course of the argument, I cited a [...]