July 3, 2012 | 118
Notice to those who see that this article is incredibly long and then decide not to read it: here’s the take-home point… ReptileEvolution.com does not represent a trustworthy source that people should consult or rely on. Students, amateur researchers and the lay public should be strongly advised to avoid or ignore it.
For the rest of you…
The study of Mesozoic archosaurs – dinosaurs and pterosaurs in particular – attracts a great many interesting people who might best be considered ‘outside’ the normal, academic community. These wonderful individuals span the range from the deranged, dangerous, unhinged ones who threaten you with physical harm, to the erudite, technically adept and sometimes likeable people who succeed in publishing some of their less problematic opinions and hypotheses in the technical literature. Sitting somewhere in the latter part of that cluster we find David Peters.
David Peters has been a constant background presence in the Mesozoic reptile research community since the early 1990s, mostly because of his interest in pterosaurs. Actually, Dave first made a name for himself as a very competent artist, and several books on living and prehistoric animals showcase his brilliant work (Giants of Land, Sea & Air, Past & Present (1986), A Gallery of Dinosaurs & Other Early Reptiles (1989), From the Beginning: the Story of Human Evolution (1991), Strange Creatures (1992), Don Lessem’s 1996 Raptors! The Nastiest Dinosaurs and Don Lessem’s 1997 Supergiants! The Biggest Dinosaurs). I and many others think very highly of Dave’s artistic skills, and I understand that he is a highly successful businessperson in the world of marketing, advertising and web design.
I began corresponding with Dave during the mid-1990s and got to see early versions of his pterosaur skeletal reconstructions, his markedly unusual phylogenetic trees for Pterosauria, and an unpublished manuscript in which he argued that pterosaurs were unlikely to be close kin of dinosaurs (it was titled “Questioning the inclusion of pterosaurs within the Ornithodira*”). It was obvious right from the start that Dave was adept at using his skills in graphic design and image interpretation to ‘see’ stuff that was generally ignored or dismissed by other researchers, and that he was unafraid of making wholly novel proposals based on these new and unusual observations. I didn’t think that his phylogenetic trees were likely to be right, but in general I thought that this stuff was great – finally, someone who could find the details that we so often want to see in the fossil animals we’re interested in.
* The title is inaccurate since, by definition, Ornithodira is the clade that includes Dinosauria and Pterosauria. Pterosaurs will always be ornithodirans, no matter where they end up in the tree of life.
A necessary disclaimer
Before I continue, I feel it’s important to get some personal stuff out of the way. The article you’re reading was difficult to write and took a long time to put together. Partly this is because I’m acutely aware of the fact that it might look like a personal attack. So I want to get some things very clear right from the start. This is not intended as a personal attack on Dave Peters.
Dave is an independent researcher, unaffiliated with any academic institution, and without qualifications in the earth or biological sciences. But… I don’t care. I’m not an academic snob; I believe that your value as a researcher is judged by what you bring to the table and on what you contribute, not by your qualifications or affiliations. Dave Peters has numerous unique, heretical views on both the life appearance of the animals he’s interested in, and on how those animals might be related. I strongly disagree with just about all of his claims and arguments (and so do other researchers who have expressed an opinion); the reasons for this are discussed in the rest of this article.
But I’m not saying that he should be censored, and I think that he has every right to share his observations and ideas. These days, the internet is the tool for that. But, despite success at publishing technical papers in mainstream and even top-tier venues, Dave has, increasingly, failed to get his heretical ideas into the technical published literature or discussed at palaeontological meetings. Is Dave’s peripheralisation a result of the fact that those of us who review technical papers and talk abstracts belong to some sort of secret club, that we’re fighting to defend a status quo, or to maintain the ‘traditional’, ‘establishment’ views that are otherwise prevalent in the community, or is it that Dave is ‘blackballed’, as he contends? No. It’s because reviewers, as experienced, sceptical scientists, can see unreliable and inaccurate work, and they act to prevent its appearance in the technical scientific arena.
Frustrated by mainstream rejection of his ideas and proposals, and incredibly productive to boot, Dave has put huge quantities of his personal research online at a new website that he’s called ReptileEvolution.com. That’s why we’re here.
ReptileEvolution.com has an enormous web presence. This presence is so pervasive that Dave’s heretical views are mis-educating naïve parties who encounter his material, and think that they’re seeing something worthy or accurate. Keeping in mind everything I’ve just said in the last few paragraphs about this not being a personal attack or attempt to censor Dave Peters or his views, those of us interested in tetrapod evolution need to counter Dave’s online work and proclaim as loudly and publicly as possible: ReptileEvolution.com does not represent a trustworthy source that people should consult or rely on.
Dave runs a second internet site – a blog called The Pterosaur Heresies, subtitled “There’s something very wrong with our pterosaurs”. I don’t agree with any of the stuff he says there, either (and, in fact, some of it I find personally insulting), but the fact that the site is labelled ‘Pterosaur Heresies’ at least lets naïve readers know that they’re going to be reading about non-mainstream views that are heretical compared to what’s in the rest of the literature.
So… I really hope readers can understand why I’m writing this article. It is not an attack on Dave Peters. It is not an effort to censor or muzzle him. Plain and simple, it is an effort to strongly express the opinion – as best and as publicly as I can – that Dave’s observations and ideas, as published online at ReptileEvolution.com, represent a highly idiosyncratic, almost certainly wholly erroneous, view of tetrapod anatomy and evolution. His stuff is definitely fun to look at, but my concern is that naïve parties are effectively being mis-educated by it. This article is an effort to counter that mis-education. Again, it is not a personal attack.
Because referring to Dave Peters as ‘Peters’ in the rest of this article could be misconstrued as sounding aggressive (even though it would be in keeping with the convention of scientific writing), I’m deliberately going to keep things informal and refer to him as ‘Dave’ or ‘David’ throughout the rest of this article. I’ve met and spoken with Dave on several occasions and would like to maintain as cordial and friendly a relationship as possible. I emailed him on June 30th to let him know that I was writing this article and we then corresponded about its contents. He very kindly gave me permission to use his illustrations. I have tried to be as fair as possible in this article, but it’s very clear that we will never agree on the key points of contention. Dave says that he will respond to this article.
The backstory to ReptileEvolution.com
The journey that ended with the production of ReptileEvolution.com can be imagined to consist of five stages. These stages chart Dave’s increasingly elaborate, increasingly unorthodox opinions and ideas. An article about the evolution of Dave’s ideas is already online: Nima’s 2011 The Strange Journey of David Peters. It catalogues many of the responses and rebuttals to Dave’s ideas that have appeared over the years. Anyway, here’s my take on the same subject…
Stage 1. Starting in 1995, Dave began publishing arguments and hypotheses in the technical literature. He started with a brief letter in Nature (Peters 1995) where he suggested that previous authors (Unwin & Bakhurina 1994) had erred in their interpretation of a particular pterosaur’s wing membranes (that particular pterosaur was Sordes pilosus, a small, long-tailed form from Kazakhstan). What happened in that 1995 article set the course for everything that was to follow: Dave looked at published photos of the pterosaur concerned, thought he could see something that the original authors (and everyone else who’d looked at the actual fossil) had missed, and based his whole argument on the re-imagining of an image (Peters 1995). In their response to Dave’s article, Unwin & Bakhurina (1995) noted that “Peters’… reconstruction… is based on a highly unreliable technique, interpretation of photographs” (p. 316).
Stage 2. Dave’s next significant article (“significant” in the context of the story I’m telling here) was very different from that half-page Nature letter; it was a lengthy article titled “A reexamination of four prolacertiforms with implications for pterosaur phylogenesis” (Peters 2000). It’s an interesting paper in purportedly providing a huge amount of new anatomical information on the frustratingly enigmatic and highly weird Triassic reptiles Longisquama and Sharovipteryx, as well as on the less enigmatic and less highly weird Cosesaurus and Langobardisaurus.
Based on his examination of these four Triassic animals, Dave used several different phylogenetic analyses to show that they were close kin of pterosaurs. If correct, this would move pterosaurs well away from dinosaurs in the reptile family tree.
I readily admit that I was initially impressed with this body of work – wow, how was Dave able to find so many brand new bits of both skeletal and soft-tissue anatomy? Longisquama is famous for being weird. It has giant, feather-shaped appendages of some sort projecting upwards from its back. Dave was able to go much further. He had reconstructed the skull in more detail than anyone had before, and apparently succeeded in finding many new details of the forelimb and pectoral girdle (Peters 2000). For the also enigmatic Sharovipteryx, he had apparently discovered many new details of the pectoral girdle and forelimb as well as innumerable, detailed features of the skull, including even the pattern of scales on the snout tip.
Excited, I spoke to others about this new paper (Peters 2000). They were far from enthusiastic; the overwhelming opinion being (1) that you just couldn’t see the things reported in that paper in real life (that is, the interpretations were in error), and (2) that the phylogenetic analyses included in the paper were completely erroneous, since they wholly relied on those problematic observations (in the paper, Dave added characters coded from his re-interpretations of Langobardisaurus, Cosesaurus, Sharovipteryx and Longisquama to the published data-matrices produced by other workers).
That 2000 paper simply did not instigate any sort of paradigm shift in mainstream thinking about the affinities of pterosaurs: other workers have continued to find support for a close affinity between pterosaurs and dinosaurs (e.g., Brusatte et al. 2010, Nesbitt 2011, Butler et al. 2011), and they’ve ignored the results and proposals of Peters (2000). Is this because Dave is somehow “outside” “establishment” thinking? Or is it because his work contradicts the personal positions of the scientists who act as reviewers of his research, as Dave has repeatedly suggested? No, absolutely not. It’s because his observations, and hence his character codings and the trees that result from them, are regarded as wholly unreliable. I say more about these issues below. Remember that the workers who produce the trees that support the ‘conventional’ view of pterosaur affinities have spent a lot of time looking at actual fossils. Furthermore, these researchers have done an excellent job of explaining and documenting the characters they use in their phylogenies, and they have proven track records of showing that they understand how to reconstruct phylogeny and analyse phylogenetic data.
Anyway – - – claiming to find those new details of Sharovipteryx and Longisquama was remarkable enough, but things were going to get much, much weirder.
Stage 3. The Longisquama holotype consists of the front half of the animal, preserved on a slab of matrix. It seems that the adjacent chunks of matrix are known, but they’ve rarely been figured in the literature. By using a special photo-tracing technique [read on] on both the front half of the specimen, and on the additional segments of matrix, Dave claimed some time round about 2003 that he (and everyone else) had previously understated the weirdness of Longisquama. He claimed to find the whole back end of the animal – the hips, the hindlimbs, the tail, and a whole bunch of additional, giant appendages. And hitherto-overlooked baby specimens of Longisquama were preserved on the slabs as well.
Dave had now crossed the line from producing remarkable-but-just-about-plausible results to freakin’-crazy-almost-certainly-not-real results. Longisquama-like dorsal frills, claimed Dave, were present in pterosaurs. Yes, in ALL pterosaurs. Short-tailed pterosaurs were not actually short-tailed: they actually, said Dave, have long, whip-like tails, often with tassels at the ends. Furthermore, toothless pterosaurs actually have teeth after all, wing-fingers still have claws at their tips, and digit V is still present in the pterosaur hand, says Dave. Some pterosaurs have two nostrils, says Dave (as in, two on each side). Dewlaps, enormous dorsal crests and even anglerfish-like ‘lures’ decorate the heads, snouts and throats of Dave’s pterosaurs. Some of these structures, says Dave, are about as big as the animal’s head and body combined. ‘Best’ of all, pterosaurs of many kinds are – says Dave – preserved with babies attached to, or adjacent to, their bodies. These babies are, says Dave, built like miniature adults (that is, they have adult-like proportions), decorated with the same grandiose flaps, frills and dewlaps as Dave’s adults, AND they have wholly unossified skeletons, hence explaining their feint, translucent, near-invisible essence on the slab. I’m not joking about any of this stuff: it’s all documented online, and also in Peters (2004).
Stage 4. Like anyone interested in pterosaurs, Dave wants to know how these animals might be related to other reptiles. As noted above, the general thinking on this issue is that pterosaurs are archosaurs, closely related to dinosaurs and their kin, and forming with them the clade Ornithodira. A list of anatomical characters appears to support this position. However, it has been suggested on occasion that pterosaurs aren’t archosaurs at all, but are in fact more closely related to a poorly known group of Triassic reptiles that are probably best termed protorosaurs (other names have been used, but things will be simpler if I avoid that mess here). This is the position supported in Peters (2000). Therein, Dave found Cosesaurus, Sharovipteryx and Longisquama to be successively closer relatives of Pterosauria, with protorosaurs like Tanystropheus and Macrocnemus being outside Fenestrasauria, the name Dave gave to his proposed Cosesaurus + Pterosauria clade (Peters 2000).
In more recent work, Dave has taken to looking in detail at other kinds of reptiles, and indeed at other kinds of tetrapods. Some time round about 2007/2008, he started saying how new work was revealing that pterosaurs (and other ‘fenestrasaurs’) were close to, if not within, Squamata (the clade that includes snakes, lizards and amphisbaenians). And this radical new proposal – that pterosaurs are not archosaurs, nor even close, but actually nested within squamates – was not alone. Dave claimed that he’d ‘discovered’ a whole new shape for the tetrapod tree with numerous unorthodox relationships. More on this below.
Stage 5. So far as I know, Dave hasn’t succeeded in getting this new hypothesis of tetrapod phylogeny into print, despite efforts (I know for a fact that he’s submitted manuscripts on this new model to certain top-tier publications) [UPDATE: not wholly correct. Brad McFeeters has brought my attention to Peters (2006), a popular magazine article in which Dave discusses his new tree]. He’s therefore taken to putting it all online. And this brings us up to the modern day. All of Dave’s observations, ideas and phylogenetic hypotheses are discussed at length at his new website, ReptileEvolution.com, and at his blog site, The Pterosaur Heresies.
Here we come to the reason for the article you’re reading now. ReptileEvolution.com is a huge problem, and those of us involved in research on tetrapod evolution need to somehow counteract its influence. If you google the name of just about any fossil reptile, synapsid or amphibian (I’m using ‘amphibian’ in the maximally inclusive, enormously vernacular and technically wrong sense), you get numerous hits for ReptileEvolution.com, typically high up or even top in the search results. This goes for image searches as well as for normal ones. And, while it’s understandable that this goes for obscure animals where there’s comparatively little information available online (I tested by googling Silvanerpeton, Kotlassia, Owenetta, Eusaurosphargis, and Helveticosaurus), it goes for comparatively well known ones too (I tested by googling Platyhystrix, Seymouria, Diadectes, Placodus and Coelurosauravus – believe it or not, these are “well known” as fossil tetrapods go).
I’ve been saying on facebook for a while (we talk quite a lot about Dave’s stuff) that something needs to be done. By this I mean that the rest of us need to somehow counteract the enormous web-presence of ReptileEvolution.com. So, what to do? Just ignoring ReptileEvolution.com doesn’t work in the age of the internet. The site is easy to find, so any rebuttal of any sort needs to be easy to find too.
The high visibility of ReptileEvolution.com is exacerbated by the fact that Dave’s ‘competitors’ – those who publish the sort of big-scale, taxon-heavy analyses that contradict his own trees – either have no internet presence at all, or are not interested in putting images, diagrams and other representations of specimens online. And image use is one of the key issues here. Dave is, as we’ve seen, a brilliant artist: he has been, and is, able to generate hundreds of excellent, highly attractive images. His stuff looks great. It’s therefore not really much of a surprise that naïve parties, after finding his stuff online, are drawn into thinking that it might be good, reliable work.
It might, by now, be obvious that I have a problem with ReptileEvolution.com, and indeed with Dave’s work in general. But we’re not done yet. Let me spell it out as clearly as possible. Why, exactly, am I so unhappy about ReptileEvolution.com?
A whole lot of heresy. Whether we like it or not, few of us will make contributions to the human endeavour that will persist far into the future. I like to imagine that the total sum of human knowledge can be imagined as a giant, sprawling, baroque castle made of lego bricks. It’s bristling with lego flying buttresses, lego gargoyles, lego battlements and lego towers, and it’s surrounded by defensive lego walls. Everything is connected, but everything is incomplete and in need of more lego. Those of us who publish scientific results will, if we’re lucky, get to add a few bricks to one of the walls, or perhaps get to slightly modify the shape of one of the battlements or buttresses. People who get to build whole new walls, or wings, or – better yet – demolish an existing wall or wing and construct a wholly new one are rare and special indeed.
What I’m getting at here is that people who come along and properly instigate paradigm shifts or convincingly overturn long-held models are exceptional, and either incredibly gifted, incredibly lucky, incredibly hard-working, or incredibly rich… or some or all of the above.
David Peters would have us believe that just about the whole ‘mainstream’, accepted structure of the tetrapod tree is wrong, and that he – uniquely – has discovered a wholly new, paradigm-busting one. According to Dave, (1) reptiles and all other amniotes can be placed along either a lizard branch, or a mammal-croc-bird branch; (2) pterosaurs are not archosaurs or even close, but deeply nested within lizards; (3) synapsids (the amniote clade that includes mammals) are on the croc-bird branch, close to (or part of) Archosauromorpha; and (4) Dinosauria does not consist of Ornithischia and Saurischia, but Theropoda and Phytodinosauria (and the latter clade includes several traditional non-dinosaurs, like Lotosaurus and silesaurids).
And there’s more – lots more. Synapsids are non-monophyletic according to Dave (since some traditional synapsids – caseasaurs – are on the lizard branch in Dave’s tree). Snakes are non-monophyletic, with scolecophidians (worm-snakes and kin) grouping with some platynotan lizards, rather than with other snakes. Trilophosaurids and rhynchosaurs are in Rhynchocephalia, close to sphenodontians. Mesosaurs are close kin of ichthyosaurs and thalattosaurs. And so on…
If you’re not familiar with the generally accepted structure of the tetrapod family tree, it would take me a while to explain why these proposals are so heretical and contrary to other studies. Suffice it to say that they are odd indeed, totally discordant with the careful, detailed work that other workers have documented in the peer-reviewed literature (for the record, a highly simplified ‘consensus’ tetrapod tree is depicted below).
The specific details of phylogenetic hypotheses – that is, the specific positions of species relative to one another – will, in many cases, always fluctuate between studies. But the general pattern of the tetrapod cladogram as a whole is universally agreed on: mammals are the sister-group to the turtle-lizard-croc-bird clade; lizards (and other lepidosaurs) are the sister-group to the croc-bird clade. This pattern has been consistently recovered in morphology-based analyses (e.g., Laurin & Reisz 1995, Müller 2003, Lee et al. 2004, Hill 2005) and is also overwhelmingly supported by molecular data (e.g., Hedges & Poling 1999, Janke et al. 2001, Rest et al. 2003, Shedlock et al. 2007, Lyson et al. 2011). Embryological, biochemical and behavioural data also shows, pretty much unanimously, that lizards form a clade with archosaurs while mammals are outside this lizard + archosaur group.
In additional to this phylogenetic re-shuffling, and in addition to those many new details of anatomy that he claims he’s discovered (more on that in a moment), Dave also thinks that he’s discovered some crucial new stuff about the biology and behaviour of pterosaurs and other fossil tetrapods. Using the digital tracing technique, he claims to have discovered flightless pterosaurs, vampiric pterosaurs that bit dinosaurs, widespread evidence of super-narrow wing membranes, and even prey items (like insects) preserved within the mouths of some animals. Pterosaurs have generally been assumed to be egg-layers, an inference based mostly on their hypothesised position among archosaurs. Recent finds of baby pterosaurs preserved within eggs (Chiappe et al. 2004, Ji et al. 2004, Wang & Zhou 2004), and of an egg preserved right next to the pelvis of a particular pterosaur specimen (Lü et al. 2011), provide compelling support for that assumption.
But Dave’s claim that numerous unossified baby pterosaurs are preserved alongside – or on or even in – the bodies of adult specimens is discordant with this, since their ‘presence’ led Dave to argue that pterosaurs were viviparous. After the first baby pterosaur preserved inside an eggshell was discovered, Dave seriously proposed that it represented a miniature kind of pterosaur – he named it Avgodectes pseudembryon – that took to hiding inside broken eggshells (he published that name in the magazine Prehistoric Times). I should add here that I don’t think that those numerous, soft-boned babies exist at all; they’re artifacts of the digital tracing technique (on which, read on), and in fact it might (or might not) be that Dave has given up on this idea in view of subsequent discoveries.
Dave thinks that a number of small pterosaur specimens – interpreted by everyone else as juveniles of Pterodactylus and other taxa – are actually miniature adults. His interpretations are dependent on his digital tracing technique, and on the incorporation of the characters he finds via digital tracing into his phylogenetic analyses. Given that he interprets these tiny animals as adults, and given that he contends that growth in pterosaurs was isometric, he proposes that the babies of these miniature pterosaurs were less than 10 mm long. Yes, less than 10 mm long.
This is an awfully long list of heresies to come from one researcher. Is it impossible that Dave Peters really is the most insightful, most gifted, most brilliant compiler and analyser of phylogenetic data of our time? No, it is not impossible. Is it likely? Let’s consider the following.
How not to change the world. If you think you’ve discovered something radically new and absolutely contradictory to previously accrued evidence, you don’t go round saying “Hey everyone – I’ve solved all your problems. Everything is different now – here is the new truth!”. You ask other people what they think before your announcement, you try and test it using alternative methods (for fossils, the application of CT-scanning and UV light to specimens is showing us stuff we can’t clearly see with our own eyes) and, essentially, you state your incredible discovery in appropriately conservative fashion. Some people actually take years or even decades to announce their major, groundbreaking discoveries because they want to be as sure as they can be that they’ve considered all the weaknesses, alternative explanations and possible flaws in what they’re proposing.
Dave is a bit of a contradiction on this front. He’s thrown a million radically strange new discoveries out there at a phenomenally rapid pace, and indeed the rate at which his ‘discoveries’ occur is unprecedented. Dave proclaims frequently that he changes his ideas when he’s wrong, and indeed he invites others to test his claims. So far so good. But, when others don’t see what he sees, when they criticise his interpretations and his methods, he remains steadfast in his opinion that they’re wrong because they’re biased, because they’re refusing to use the same method that he does (read on), or because they can’t provide a superior hypothesis.
What about the alternative – that they’re not wrong? I’ve now corresponded with Dave on several occasions about the structures he reports to find. He seems very confident that he’s always right, yet I don’t think that he ever is. I am not alone; many others have challenged Dave’s observations in discussion (virtually all of this is online, though see Bennett 2005), and indeed Dave’s work is ignored by publishing scientists. And so we come to…
The core problem: Digital Graphic Segregation. We convince others of our observations by pointing to objects or structures that can be perceived by those others. If I assert that structure x is present in specimen y, I say so with the confidence that any other sighted person will be able to look at that specimen (or an image of it) and, like me, see the same structure. Seeing as a lot of my technical work has involved descriptive skeletal anatomy, I think about this issue quite a lot; I think that any ambiguity surrounding the existence of anatomical structures should be avoided: something is either demonstrably there, or it’s not. Today, we can confirm the existence of ambiguous or hidden elements by using new imaging technology, like the CT-scanning and use of UV light that I’ve already mentioned.
Contrast this with the technique that makes Dave’s observations rotten to the core. He calls it Digital Graphic Segregation or DGS (so far as I can tell, this term is unique to Dave’s writings), and he claims that it allows him to, variously, separate bones from underlying ones, to piece together previously unnoticed, shattered bits of elements in jigsaw-fashion, and to recognise truly novel, soft-tissue structures like elaborate dermal frills and those unossified pterosaur babies.
David maintains that, via DGS, he can reliably distinguish and identify tens and sometimes hundreds of bones and other structures that are effectively invisible to everyone else. A primary assumption behind the technique is that fossil animals preserved flat on slabs of sediment are always preserved as ‘high-resolution, high-fidelity’ objects, where even near-microscopically tiny elements are preserved, findable, observable, and decipherable. As you’ll know if you’ve spent time looking at fossil vertebrates preserved on slabs, this is very typically not the case – many fossils are preserved as partially eroded, heavily broken, very much incomplete objects where tiny elements are invisible, long gone, or unpreserved, and where eroded, degraded and distorted bone surfaces are wholly lacking their sutures and many of the fine details they would have had when fresh. They are just not the ‘high-resolution, high-fidelity’ objects that Dave assumes.
I have no doubt whatsoever that, with literally one or two exceptions, every single element Dave is identifying via DGS can be explained in one of three ways. (1) It’s not a genuine anatomical structure or feature: it’s a scratch or imperfection on the slab, a lump or bump in the adjacent sediment, an area of paint, glue, preservative or cement, a bit of plant tissue, or even a digital artifact resulting from the coarse-grained pixelisation that occurs when you zoom in close to scanned images (for this work, Dave relies almost wholly on images from the published literature). (2) Some of the big, ragged structures that Dave interprets as frills, dewlaps, wattles, pouches or crests might actually be sloughed patches of tissue that have broken away from the animal’s corpse during decomposition. Dave’s default assumption for such structures is that they can be interpreted as being preserved in the ‘in life’ position, but other options need to be considered first. Decomposition and fossilisation are messy. (3) In a few cases, Dave is seeing real bones, but interpreting them as something else. Broken, indeterminate rod-like elements that might be bits of ribs or gastralia, for example, become reconstructed as if we can be sure what their identity is.
Dave says that DGS is reliable because it revealed one unquestionably valid, genuine discovery: he found the head and neck of Langobardisaurus pandolfi thanks to this method. But this is a red herring. Look at the fossil (it’s here; below). I maintain that you do not need some special digital tracing trick to see that missing head and neck: you can clearly see the skull preserved, upside down, beneath the ribcage. Compare this with the other fossils where Dave claims to make ‘discoveries’. They are not the same.
Every single researcher who has expressed an opinion on this issue has said exactly the same thing, and you can see from discussions in chatrooms, message boards and blogs that interested amateurs, fanboys/girls and idly curious commenters tend to note that Dave’s views and DGS ‘discoveries’ are suspicious or obviously erroneous too. That is, Dave is ‘discovering’ artifacts of sedimentology and taphonomy, not genuine features of anatomy. Chris Bennett published an excellent article on Dave’s methods and observations in which he pointed to specific case studies (Bennett 2005). That article is required reading for anyone interested in the Dave Peters issue [a pdf is linked to in the references below].
In email correspondence, Dave said that my explaining away of his ‘identified’ elements is unsatisfactory since, in order to knock down his proposed DGS-based identifications, I need to come up with superior hypotheses that are supported by “a ton of evidence”. One specific example we discussed concerns the alleged prepubes of Cosesaurus. Prepubes (singular: prepubis) are fan-shaped bones located anterior to the pelvis, and they’re unique to pterosaurs. Dave says that he’s found them in Cosesaurus, and here [above] is the photo that’s supposed to support that claim. I cannot see any prepubes there. I can see some rod-shaped impressions in about the right region, but nothing that comes close to being convincing. I thus simply reject Dave’s suggestion that he has ‘found’ prepubes in Cosesaurus. Dave is wrong in arguing that I need to present a superior explanation that accounts for the presence of the structures we can see here, or that I need to compile some sort of evidence that overturns his hypothesis. I don’t: it is wholly acceptable in science to say “I don’t know what’s going on there; we cannot reach a conclusion without more evidence”.
Dave’s mantra is “Test. Test. And test again”, his implication being that those of us who reject his proposed observations aren’t seeing the things he reports because we’re not trying hard enough: that is, we’re not testing the anatomical structure of fossils in the same way that he is. What he fails to credit is that those of us interested in fossil tetrapods are, in fact, testing his proposals every time we look at the fossils concerned. I look at the images (or at a replica, or at the actual specimen) of Cosesaurus, for example, and I make my own decision about the structures that Dave is purporting to identify. I therefore have tested his hypothesis, and I’ve rejected it. Note that future discoveries could, hypothetically, prove David right. However, we don’t work on the basis of what might be found in the future – we base conclusions on what we have now.
Dave has now made hundreds of claims concerning the alleged presence of elements and soft tissue structures ‘discovered’ via DGS, and a huge number of the reconstructions viewable on ReptileEvolution.com incorporate information obtained via this method. I think there might be – literally – one or two cases where he’s on to something. But, overall, I regard DGS as absolutely unreliable and absolutely misleading. Dave is reporting things that do not represent genuine details of anatomy. This goes for skeletal elements, dermal and soft-tissue structures, and sutures between bones. I have yet to hear from a single researcher who agrees with any of his many, many proposals. Remember that this affects all of his work, since his character codings, and hence his phylogenetic hypotheses, are heavily dependent on information gleaned from DGS. On that note…
Unreliable data sets, broken cladograms. Dave is proud of the fact that he’s created a tetrapod super-tree incorporating more species-level taxa than ever before, and where similar-looking animals group together in what he says is a thoroughly sensible, evolutionarily ‘logical’ fashion. Indeed, both ReptileEvolution.com and The Pterosaur Heresies are built around discussion of this tree.
The aim to be appropriately comprehensive in terms of coding species-level taxa is a noble one, and Dave is right to criticise previous analyses for not including enough of these, and for making assumptions about the monophyly and/or homogeneity of many groups. So, well done Dave: credit where it’s due. But Dave’s tree – the one discussed at length at both ReptileEvolution.com and The Pterosaur Heresies – is fraught with problems, and it cannot be considered a superior alternative to what’s already in the literature.
Dave challenges those who disbelieve his results to point to the problem areas – to say which taxa don’t sit right in his phylogeny. Actually, as should be clear by now, I don’t think that any of the heretical, novel positions he recovers for tetrapod taxa are reliable. I don’t think pterosaurs are close to Huehuecuetzpalli* or to other squamates as Dave contends, I don’t think that caseasaurs are close to any of the taxa that Dave nests them with (he places them well away from other synapsids), I don’t think that silesaurids and Lotosaurus are close to ornithischian dinosaurs, and so on and on. I’m not disputing Dave’s heretical proposals because I’m biased, blinkered by mainstream authority, or automatically unwilling to consider his ideas. Rather, I dispute them because (1) I find the ‘conventional’ positions to be far more convincingly supported by character data, and (2) because Dave’s alternative positions are based on erroneous or incomplete analysis of the data available.
* It’s pronounced something like ‘whey-whey-ketz-paa-lee’. I know some people who hate the name and some who love it.
As mentioned already, one huge problem with Dave’s analyses is that many of his character codings are derived from his application of the DGS technique. I’m sure that DGS is giving wholly erroneous results; all of the codings thus derived from this method are inaccurate reflections of real anatomy.
But, not all of Dave’s character codings are derived from DGS. He only uses DGS on flattened specimens that are preserved on slabs. What about the characters he uses for three-dimensional specimens? The problem is that Dave uses a DGS-like technique on these as well, re-identifying skeletal elements, re-identifying sutures and re-imagining proportions and shapes according to what he thinks is most likely. While there are cases where his reconstructions are not problematic, there are plenty of others where they are, and where his reconstructions are unreliable and almost certainly erroneous.
Furthermore, his character sets are not comprehensive enough. The tree at ReptileEvolution.com and The Pterosaur Heresies incorporates data from something like 230 characters, coded for over 300 taxa. I don’t know if that sounds like a lot to you, but I assure you that it isn’t, especially in an analysis that’s meant to incorporate data from tens of distinct clades that haven’t previously been included in the same analysis together. In recent months, I and colleagues have been working with a dataset designed for one specific archosaur clade, and it contains over 800 characters. I consider it a joke that someone can claim to comprehensively sample all synapsids, lepidosaurs, archosaurs, marine reptiles and other tetrapods and do so with c. 230 characters. Tetrapods are complicated objects, even when you’re dealing with skeletal anatomy alone, and c. 230 characters for over 300 taxa cannot be considered adequate in term of providing enough data for the recovery of a reliable tree.
When challenged to change things and add more characters and revise the old ones, Dave’s response (as of June 17th 2012) was “As long as I can recover a single tree from all these reptiles, I’m not going to add another character. Sorry. I only take things this far. No further”. I’m deeply perplexed by this statement. I honestly thought (as indicated by other statements about Dave in this article) that Dave was playing this game in order to find conclusions that best fit the data, not to give up and refuse to go further as soon as appropriate tweaking of his analysis is suggested. What the hell happened to “Test. Test. And test again”?
A great many of Dave’s characters are derived from the existing literature. That’s fine; we all look at characters that others have used beforehand, and use those that we think are useful in term of grouping taxa together. One issue: there is substantial overlap between the characters Dave uses, with many describing the same character states! (see Mickey Mortimer’s comment here at The Pterosaur Heresies). Another issue: Dave’s characters are often problematic in terms of carving up different character states, and seem cherry-picked and extremely selective. He champions heretical ideas like (to choose one example) the non-monophyly of Saurischia, yet his dataset doesn’t include half the characters that exist to support the monophyly of Saurischia in the first place. Mickey has recently been going through Dave’s characters and character codings, and he discusses some of these character-use problems in his two articles Why David Peters’ analysis sucks and Why doesn’t Peters find Dinosauria or Saurischia?.
Finally, those of us who correspond with Dave and respond to his comments on blogs and so on have learnt that he doesn’t really seem to understand how to ‘read’ cladograms, or how the results of phylogenetic analyses are supposed to be interpreted. Dave regards it as a key ‘plus point’ of his new big analysis that he only recovers one or two most parsimonious trees (MPTs), as if this is a good thing. It isn’t – the number of MPTs you recover is merely a result of how the program explains your data, and there might, or might not be, innumerable equally ‘good’ trees. Support statistics are the important part when it comes to working out the robustness of a tree; the number of MPTs itself is not an important part of tree goodness-of-fit.
Dave also consistently misunderstands the structure of cladograms, using ‘sister-taxa’ and ‘sisters’ in incorrect fashion. You might not think that this matters, but it does. Look at the cladogram below (if your memory is good, you’ll remember it from last time. It’s a tongue-in-cheek reference to the idea that crown-archosaurs are just better than other tetrapods). Dave has said several times that cladograms like this reveal pterosaurs to be the sisters to phytosaurs. But, as I hope you can clearly see, that isn’t accurate at all. In this cladogram, phytosaurs are merely one branch among several in a clade that is sister to ANOTHER clade, of which pterosaurs are also merely one branch among several. Pterosaurs and phytosaurs are, in this cladogram, nothing like sisters. Theropods (they include birds: the cladogram shown here is misleading on that point) and sauropodomorphs are sisters.
The reason this matters is that Dave – by wrongly thinking that pterosaurs and phytosaurs are ‘sisters’ – imagines that pterosaurs, according to this tree, must have evolved from a phytosaur-like ancestor. But that isn’t what a sister-group relationship means. It means that two lineages have diverged from a single ancestor. Maybe this ancestor looked more like the members of one of those lineages than the other, but maybe it looked like a combination of both, and maybe it looked like neither. In the cladogram shown here, pterosaurs have a sister-group relationship with the silesaurid + dinosaur clade (properly named Dinosauriformes). The hypothetical common ancestor would therefore have shared the characters common to pterosaurs and dinosauriforms, but lacked the special features of either of those lineages. Understanding cladograms is important if you’re thinking about ancestral character states or indeed the general appearances of concestors.
Two last things
We’re now approaching the end of this overly-long article. Well done if you made it this far. There are two more things I want to say.
I said way, way back at the start of this article that I have honestly tried very hard to not make this article a personal attack on Dave Peters. But I’ve also noted above that Dave is sometimes insulting to other researchers. The by-line to The Pterosaur Heresies is ‘There’s something very wrong with our pterosaurs’. The not-so-concealed implication here is that everyone else who works on the interpretation of pterosaurs and their kin is wrong. And Dave sometimes says stuff like the following (in this case [from this article at The Pterosaur Heresies blog], with reference to an in-press paper recently released online about the evolution of the pterosaurian pelvis)…
Why there’s something very wrong with our pterosaurs.
Giving credit where credit is due, the following pterosaur experts all had a hand in the report by Hyder et al. (2012) according to the acknowledgements: Darren Naish, Steve Vidovic, Dave Unwin, Dino Frey, Ross Elgin, Lorna Steel, Mike Habib, David Hone and Michael Benton.
In other words, Dave’s complaints about the paper concerned are directly linked to the fact that a bunch of academic researchers who work on pterosaurs – myself included – are the root cause of the problem. While no-one will deny that we all make mistakes, that there are often weaknesses in our own published conclusions and hypotheses, and that we all fail, on occasion, to understand the stuff that we claim expertise in, stunts like this are rude and unnecessary. On a similar note, Dave’s writings are consistently worded as if he’s discovered ‘the truth’, and that all that’s left is for the rest of us to investigate his data and hypotheses before we become turned around and see things his way. Please see the above section of text, titled ‘How not to change the world’.
On to my final point. Dave Peters is brilliant. He is smart, he has a phenomenal eye for detail, he is passionate, enthusiastic and hard-working, he is almost unparalleled in his ability to depict his ideas in graphical form, and his quest to reconstruct both the tree of life and the biology and behaviour of the animals he’s interested in is noble and worthy of respect. But, for all the reasons discussed above, he has gone off at a tangent and his work as it stands at the moment will never be accepted, or indeed be of substantive interest, to anyone else who has a strong technical interest in the evolution of pterosaurs, squamates, synapsids or other tetrapods. The Digital Graphic Segregation technique, the finding of innumerable things that no-one thinks are real, the radically weird, discordant-with-all-other-data tetrapod phylogeny, the problems of using too few characters, of selective character choice and other methodological problems, and the insistence that only he has seen the light and that we are all too blinkered and too biased to realise the veracity of his results effectively ensure that he works on an ‘academic island’, separated and essentially ignored by the rest of the community.
As I have – I hope – made clear throughout this article, responsible researchers elsewhere in the tetrapod research community need to stand up and help spread the word: ReptileEvolution.com cannot be used as a reliable source on any matters of tetrapod evolution, biology or behaviour. In his published criticism of Dave’s DGS-based ideas about pterosaur appearance, Bennett (2005) said that “Peters has ignored the framework of palaeontological research, has made no attempt to convince academic palaeontologists of his findings, and has taken his ideas directly to the general public and presented them as true and correct” (p. 21). Now, with ReptileEvolution.com on the scene, matters are worse.
In a recent email where I criticised his conclusions and methods, Dave said to me that I need to “get back on the right track”. I find this ironic, in that this bit of advice needs to be turned around and directed at him. I wanted to end this article by advising Dave Peters that he needs to change direction, but I can’t see that this will be received well.
In the end, Dave can continue as he pleases. The important thing – the whole reason I chose to write this article in the first place – is that the mis-education of others needs to be kept to an absolute minimum.
Refs – -
Brusatte, S. L., Benton, M. J., Desojo J. B., & Langer, M. C. 2010. The higher-level phylogeny of Archosauria. Journal of Systematic Palaeontology 8, 3-47.
Butler, R. J., Brusatte, S. L., Reich, M., Nesbitt, S. J., Schoch, R. R. & Hornung, J. J. 2011. The sail-backed reptile Ctenosauriscus from the Latest Early Triassic of Germany and the timing and biogeography of the early archosaur radiation. PLoS ONE 6(10): e25693. doi:10.1371/journal.pone.0025693
Chiappe, L., Codorniú, L., Grellet-Tinner, G. & Rivarola, D. 2004. Argentinian unhatched pterosaur fossil. Nature 432, 571-572.
Hedges, S. B. & Poling, L. L. 1999. A molecular phylogeny of reptiles. Science 283, 998-1001.
Hill, R. V. 2005. Integration of morphological data sets for phylogenetic analysis of Amniota: the importance of integumentary characters and increased taxonomic sampling. Systematic Biology 54, 530-547.
Janke, A., Erpenbeck, D., Nilsson, M. & Arnason, U. 2001. The mitochondrial genomes of the iguana (Iguana iguana) and the caiman (Caiman crocodylus): implications for amniote phylogeny. Proceedings of the Royal Society of London B 268, 623-631.
Ji, Q., Ji, S.-A., Cheng, Y.-N., You, H.-L., Lü, J.-C., Liu, Y.-Q. & Yuan, C.-X. 2004. Pterosaur egg with a leathery shell. Nature 432, 572.
Laurin, M. & Reisz, R. R. 1995. A reevaluation of early amniote phylogeny. Zoological Journal of the Linnean Society 113, 165-223.
Lü, J., Unwin, D. M., Deeming, D. C., Jin, X., Liu, Y., & Ji, Q. 2011. An egg-adult association, gender, and reproduction in pterosaurs. Science 331, 321-324.
Lee, M. S. Y., Reeder, T. W., Slowinski, J. B. & Lawson, R. 2004. Resolving reptile relationships. In Cracraft, J. and Donoghue, M. (eds), Assembling the Tree of Life. Oxford University Press (Oxford), pp. 451-467.
Müller, J. 2003. Early loss and multiple return of the lower temporal arcade in diapsid reptiles. Naturwissenschaften 90, 473-476.
Nesbitt, S. J. 2011. The early evolution of archosaurs: relationships and the origin of major clades. Bulletin of the American Museum of Natural History 352, 1-292.
Peters, D. 1995. Wing shape in pterosaurs. Nature 374, 315-316.
- . 2000. A reexamination of four prolacertiforms with implications for pterosaur phylogenesis. Rivista Italiana di Paleontologia e Stratigrafia 106, 293-336.
- . 2002. A new model for the evolution of the pterosaur wing – with a twist. Historical Biology 15, 277-301.
- . 2004. Pterosaurs from another angle. Prehistoric Times 64, 34-46.
- . 2006. The reptile test. Prehistoric Times 80, 27-29.
Rest, J. S., Ast, J. C., Austin, C. C., Waddell, P. J., Tibbetts, E. A., Hay, J. M. & Mindell, D. P. 2003. Molecular systematics of primary reptilian lineages and the tuatara mitochondrial genome. Molecular Phylogenetics and Evolution 29, 289-297.
Reynoso, V.-H. 1998. Huehuecuetzpalli mixtecus gen. et sp. nov: a basal squamate (Reptilia) from the Early Cretaceous of Tepexi de Rodríguez, Central México. Philosophical Transactions of the Royal Society of London B 353, 477-500.
Shedlock, A. M., Botka, C. W., Zhao, S., Shetty, J., Zhang, T., Liu, J. S., Deschavanne, P. J. & Edwards, S. V. 2007. Phylogenetics of nonavian reptiles and the structure of the ancestral amniote genome. Proceedings of the National Academy of Science 104, 2767-2772.
Unwin, D. M. & Bakhurina, N. N. (1994). Sordes pilosus and the nature of the pterosaur flight apparatus Nature, 371, 62-64 DOI: 10.1038/371062a0
- . & Bakhurina, N. N. 1995. Wing shape in pterosaurs. Nature 374, 316.
Wang, X. & Zhou, Z. 2004. Pterosaur embryo from the Early Cretaceous. Nature 429, 621.