This article cropped up online this week and has been followed up with various online discussions about the legitimacy of at least some Chinese fossils, and with Tianyulong being a big surprise and the unfortunate history of ‘Archaeoraptor’* there have been questions asked about how we can tell if these are real or not. Some of the discussion has been helpful, but I think much of it has been based on unrealistic expectations of researchers and museums, and a misunderstanding of how you can tell fakes from non-fakes apart. Given my knowledge of Chinese material and work with Helmut Tischlinger on UV lighting I thought I should probably pitch in on this and try and add some more information.
First off, are there fakes in Chinese museums?
Yes, certainly. But then there are probably fakes or chimeras in a great many museums that have been described at some point and not all of them have been found, or will ever be. Science is self correcting and we are still finding fakes and chimeras in our collections but no-one seems to be too worried about that. There are plenty of fossils that have at least been ‘tweaked’ at some point (the ‘Zittel wing’ springs to mind) to make them ‘better’. Provided you can identify them, it’s not really a problem, you just have to be careful in what you say about them. Don’t forget that half of ‘Archaeoraptor’ was a new species that was subsequently described. That’s not to say China does not have a problem, but it is the only one currently taking flack for this and it’s hardly unique in that respect.
Next, how are these fakes getting into collections?
Well first of all we have to decide what a scientific collection is. There are plenty of museums in China that are privately owned, or are callem ‘museums’ but act as shops for selling fossils (generally illegally), or have little or no staff with any scientific training. As a result they buy things that look ‘good’ even if most researchers would spot them as a fake almost instantly. Many institutes probably have no fakes at all – I have yet to see anything in the IVPP that comes under any suspicion (except something we deliberately acquired as a fake) though we collect most of our stuff from the field in any case which obviously eliminates this issue entirely. It is typically (if not exclusively) the fossil dealers who trade fakes, or the farmers who ‘improve’ their material to fetch a higher price.
So how are the fossils faked?
There are two different things going on here, first of all you can see people just carving fake skeletons (or far more often eggs) out of rocks. The second option (as seen with ‘Archaeoraptor’) is to create chimeras by stitching together several specimens to make one ‘good’ one.
How can we detect these fakes?
The carved skeletal ones are usually are pretty easy to spot as there is no difference in the geology of the matrix and the fossil, they lack anatomical details, are carved out of the wrong rock types, contain carving errors and more. The eggs however can be incredibly good with all kinds of details and tricks employed to make them more believable. A small sample taken from either can also conclusively show that it is made of fossilised organic material or otherwise.
Testing the chimeras is another problem entirely. Most are still obvious – I saw a pterosaur not long ago that had the metacarpal block installed as a fib-tib complex and also featured three femora – two in the thighs and a third replacing a humerus, and one wing had 5 phalanges and the other 3. Despite the testing ‘Archaeoraptor’ underwent, it was obvious that the hindlegs were mirror images and a result of a single fragmented leg being used to provide a missing one, and the variation in preservation of the feathers, bones and even the underlying matrix looked wrong. Suspician was raised about it by researchers just from early photos without even needing to get their hands on the actual specimen to examine it in detail.
There are plenty of clues to look for that either indicate a fake, or are sufficient to raise questions about it where it can be examined in much more detail: exceptionally poor preservation, inconsistent rock or bone colours, non-matching parts, variation in preservation, bones or other features not crossing breaks, patterns of ontogentic changes (like the sequence of suture fusion), obvious mismatching of bones (like a pterosaur bone in a theropod body), orientations of the grain of the rock or bone being inconsistent, over zealous use of glue to replace fragments, unlikely claims of providence or discovery and others, all of which can help identify fakes or partial fakes. This may not necessarily make for an absolute diagnosis, but it should certainly reveal which specimens are likely to be problematic and require further investigation, or more careful examination.
Several people have pointed to UV light or X-rays as the solution but this is largely impractical. First of all one does not simply ‘shine UV light’ on the specimen to revela it in all it’s glory. It requires lights of the right wavelength and power and skilled use of filters to show off organic remains. Certainly the bones are likely to fluoresce, but that does not tell you if you have a chimera or not. It *can* show where a fake part has been carved or painted onto a fossil as it will be disjunct in reflectance, but it cannot tell apart one ‘real’ part form another unless the preservation or taphonomy is dramatically different (and that’s probably clear under normal light) and that is typically not the case in the Jehol beds. Obviously our work remains unpublished, but the studies Helmut Tischlinger did here in Beijing on Jehol material suggests that there are at least on occasion remarkable variations in preservation of both bones and soft tissues under UV light even in single specimens or between plates and counter-plates. As a result, even a poorly made chimera would not necessarily be detectable under UV light as it could look like an unusual, but ‘real’ specimen. It’s also impractical for museums or teams of researchers to buy expensive equipment and train people to use it all over China, certainly in the short term. There are very, very few people with this kind of expertise and equipement in the world, and while I know of other teams experiemnting with UV, Helmut is the only person I know of doing extensive work in this field.
X-ray, MRI and CAT scans of specimens can reveal differences in preservation or origination of parts of specimens but these are also impractical. These are expensive to use and rare in China – it’s not like there are several sitting in each small town or city, and most are medical scanners largely unsuited to working on rocks. One simply cannot transport very valuable and often very fragile fossils over huge distance to stick in machines at great cost that may not reveal anything in any case. This situation will improve in the future as funds are made available and more machines are present and owners might be more willing to transport their specimens, but is not practical or even possible at the moment for many fossils.
Ultimately the only way to tell at least some may be to take multiple bone and rock samples of the questionable parts or specimens and test them. Isotopic dating can see if they are all the same age, chemical tests can ensure that all the parts of the matrix are from the same beds (though a fake of two specimens from the same horizon would still show up the same). Examining sections of the bones could tell if all parts of the skeleton were of the same age (was the femur 5 years old and the tibia only 2). That of course is potentially very damaging to the specimen, costly in terms of time and effort, and not even always practical (most specimens are squashed flat or broken so you simply can’t section them).
Despite this pessimistic line I have taken here, in general these things are not hard to spot. The fact that researchers talk about spotting the fakes reveals that they can find them with often only cursory looks. With detailed examination with a lens or microscope when you are specifically checking for faked or reconstructed parts and any signs of them it is better still. One must remember that the people making these things are typically uneducated farmers or fossil dealers. They don’t realise that we have the anatomical and geological knowledge to spot changes that they cannot or do not notice. We can tell when a femur is in the wrong orientation, or when a humerus from another animal has been put in, or that the sacrum has too many vertebrae, that the head on of a different type of rock to the feet, or the bedding planes run in different directions between the left and right sides. The sheer poor quality of some is almost hilarious – animals with three legs, or the head put at the end of the tail. Palaeontologists don’t buy or study these kinds of fakes. Anything so well made and put together it can’t be spotted short of multiple bone samples is probably composed of two specimens of the same species form the same horizon of the same quarry and would be indistinguishable to anyone and would be hardly different to many other probable chimeras in collections worldwide (if you find non-overlapping parts of a skeleton in a quarry that is all of one species you will probably treat is as a single specimen even when it may not be).
The reality is that it comes down to not buying fossils of uncertain provenance, and a good eye for basic anatomical and taphonomic knowledge and experience to spot fakes. The former can be applied through eliminating the illegal trade in fossils, the former through education and training.
There are fakes in Chinese museums. But we must distinguish between research institutes with scientific aims and methods which house few, if any, fakes and those which are vanity collections or of fossil dealers which may be littered with them or even producing them. Do not tar Chinese collections as ‘full of fakes’, nor assume that fakes are either incredibly common or hard to spot – in general they are neither. Mainstream researchers are neither acquiring fakes nor working on them, and are actively looking out for those that might be and have the experience and the knowledge to spot them.
* For those unfamiliar with the story of ‘Archaeoraptor’, in short this was a faked chimera that was presented as a real fossil and ‘published’ in Natural Geographic, but was exposed as a fraud before it was described scientifically in a proper journal.
Dave Hone's Archosaur Musings 
One of the more common type of fakes is when incomplete specimens are reconstructed with materials such as plaster and then described in the literature as if it were real bone. I once was handed the “squamosal” of the holotype of a well known Triassic archosaur that was completely made of chickenwire and plaster and then painted. Unfortunately I misplaced it!
Sometimes this was done in the “old days” not out of maliciousness, but rather to have good exhibit specimens. Sometimes the reconstruction is so good it is impossible to determine what is real without dismantling the specimen. Interestingly this also happens in architecture. Conservationatists take great pains now to make reconstructions in historic buildings slightly different from the original so that modern repairs can be readily discerned by experts.
My father was not a paleontologist but an English literature and library service academic, specializing in “descriptive bibliography.” He had occasion to study a certain number of forgeries. (People pay money for dinosaur fossils and scientists get famous for discovering/studying them. People pay money for rare books and scholars get famous for finding/studying them. Need I say more?)
His considered opinion was that at any point in time it is possible to produce a forgery that is undetectable, but that at later times (when “science”/technology have advanced– in his field, the science included detailed history of printing and paper making) they would all be detectable.
(((For recreational reading, try to find out about the “Wise Forgeries” of Thomas J. Wise: there is a fun book, “Forging Ahead,” about them, by, I think, Richard Altick.)))
Bill, that is a good point. We certainly have ‘repaired’ specimens here in the IVPP in large numbers (as do a great many museums worldwide) and in that case not only is the plasterwork often easily spotted, but of course reference to the original description and images usually make it quite clear what is ‘real’ (all bone) what is ‘modelled’ (all plaster – as opposed to faked) and what has been ‘repaired’ (bone & plaster). Gasosaurs which i am working on here is a fine example, and I have documented on the Musings the troubles I went to to try and clean off the junk. As you say though, this is ‘restoration’ rahter than deliberate attempts to mislead researchers.
Allen, another good point. One exemplified by soem UV work Helmut has done where an apparently perfect Rhamphorhynchus wing turned out to be half paint. It really is all but indistinguisable under normal conditions (a microscope soon reveals it as well I should add) but is clear as day under UV. Not that we should not make efforts to find and expel or disentanlge forgeries and chimeras, but we can gaurantee that science methods and technology will improve the situation to our advantage in the future.
Great post, well written! I’ll link it on my recent post about this topic.
Are you studying Gasosaurus? God bless you!
Hi Andrea, glad you like it. Gasosaurs is at an incredibly early stage, we are still working on cleaning and sorting out the skeleton and there are, sadly, plenty of things higher up on my priortiy list but yes, technically I am worrking on it with a view to a new description.
A question: I’ve included Gasosaurus in my large-scale analysis of theropods based on Holtz’s (2000) codings and on the few informations from the original description.
Are these informations obsolete or mostly incorrect, and so, is it better to prune Gasosaurus from any analysis pending your future publication?
Well I don’t know what Tom based *his* coding on, I assume the original description, though he may have seen the material first hand – you’d have to ask him. That in itself is pretty short with few illustrations. Frankly I am still in the process of finding our how much of the material is real and how much is plaster under the ‘repairs’ done for displaying the specimen let alone cross-referencing the description with what I have in front of me. Given that it is fairly incomplete and there is a limit to what you can get from a short description in Chinese I would leave it out. I imagine you are going to have a lot of ?s, or Xs, or polymorphic codings in there and it’ll just add noise to the dataset. You can try coding it and seeing what happens with and without it, and of course STR may be an option. There is rarely any harm in at least sticking it in and seeing what happens, but at the least I can say there are more details to come out in the future. Of course you might have to wait quite a while for them though.
Given that more than 85% of my Gasosaurus OTU is made by question marks… and the limits of the original description, I’ll wait for your study before re-including it in the analysis.
(I hope to see it soon =2010?).
Thank you
Just to let you know: most of my codings for Gaso were from the original descriptions and from a specimen that was touring on display in the 1990s. That latter was clearly mostly plaster, but I tried to match up elements from the original description with the mount in choosing what was real and what was extrapolation: hence, only about 15% of the characters could be coded.
All that being said, I greatly await the redescription as a chance to toss out all the old codings and start from scratch!!
Hi Tom, good to know, I had always wondered. Yes, that is the skeleton I am trying to clean up – it’s hard ot spot some of the morphologies even now and match parts to Dong’s description (and with a student I am working on an English translation of that too to help out further). As a result (with no disrespect intended) I think any coding of Gasosaurs is going to be quite dodgy (though I think everyone knows that). I’m impressed you got as much as 15% done really.
Andrea: sounds the most practical to me. Obviously it’s good to be comprehensive, but equally it’s not alwasy a good idea to add in very fragmentary or poorly coded taxa as it can jsut add uncertainty to an analysis.
I’ve noted that the inclusion of a large number of taxa (even fragmentary) can produce an analysis more robust than one based on a smaller set of OTUs. It happens in my analysis: the a-posteriori pruning of the “unstable” forms produce a well supported tree, without the need to a-priori deletion of taxa.
I considered the position of Gasosaurus in my analysis as preliminary (pending new studies as your), mainly because I based several character codings, those absent in the phylogenies of Harris (1998) and Holtz (2000) phylogenies, only from the original description.
I hope someone will do a study similar to your on other Mid-Jurassic theropods (for example, Xuanhanosaurus).
Well two quick things, 1. a redescription of Xuanhanosaurus is already under way (not by me). 2. Yes, read my paper on the rhynchosaur Fodonyx, Mike Benton and I specifically disucss the issues of missing data and how they affect phylogenies with the situation yopu describe (phylogenies get more robust with more missing data, or less robust with more data). It’s not always the case, but can happen.
Thank you, David!
Could you give me a link or send the pdf for Fodonyx?
Funny, I was talking about precisely this subject this afternoon at the museum!
The issue as far as I see it, is that we had a specimen come from a collection which is known to contain many fakes, yet apparently it was not tested at all. This might be acceptable for low-impact journals, or low-impact specimens, but this specimen was neither.
At a very minimum, if a specimen has been purchased, it should be openly stated as such, and the person from whom the specimen was purchased from should be made clear. See the details provided for the Thermopolis Archaeopteryx for example.
Science is supposed to be transparent.
Andrea the PDF is freely available from Mike Benton’s webiste that I linked to the other day. If you can’t find it, let me know – I don’t have your e-mail address currently so I can’t send it right now.
Denver, sorry I have only just found your comment buried in the ‘waiting for approval’ section hence the late reply and late posting.
I do agree that communication here is key and that purchased fossils need to be declared as such clearly. Ideally we need to include information about *who* they were purchased from in papers or supplemental information and in collections – if we as researchers can communicate more fully on this front we can rapidly establish who has fakes and who does not. For better or worse, purchasing fossils is going to remain a part of palaeontology for a long time to come (I would argue generally worse) so we need to be clearer on who we are dealing with and what we are getting from where. We can also make it clear that we are not interested in the ‘best’ fossils so much as the ones that are of interest to us – rejecting fakes and making it obvious to those who would make them that a) we can spot them and b) we don’t want them. This information also has to go to the ‘middlemen’ and dealers who will buy fakes and then sell them onto interested parties – they generate the market (I think) as it is the non-experts buying and collecting fossils (often illegally) who produce the market for paying high prices for manipulated material as it looks ‘better’.
Wow. I’m a teenager who has been planning on becoming a paleontologist since I was three, and I found this article really helpful. Of course, my mother isn’t letting me test any of my fossils; even without the right equipment she thinks I’ll find a way. I probably can, too.
But that’s not the point. My point is, that a young paleontologist like myself could read this article and gain valuable information of fakes and chimeras.
I probably sound like an idiot, so I guess I’ll leave it at that.
—Sira
Hi Sira, no, you don’t sound silly at all. I am glad you are getting something out of this, it is why I am others write these kinds of things. Feel free to ask questions or ask for help, thought the boards on my website AskABiologist.org.uk might be more handy – there are lots of professional palaeontologist there and the site is devoted to answering questions of curious people like yourself. If you want to learn, asking questions is a good start. Listen and engage with people and it can really help you. Good luck with your plans.
Thanks! I’ll check the site out!
Hey, Dave. I just checked out Mike Benton’s website and downloaded a a bunch of awesome papers, but I didn’t see Fodonyx on there. At least, not as a PDF.
It’s available here Zach (and anyone else as well, obviously). The second paper from 2008.
http://palaeo.gly.bris.ac.uk/Benton/reprints/default.html
Heeeyyy…that’s a different page than the one I went to. Thanks, Dave.
WHAT MOTIVATED THE FAKERS?
Has anyone ever mistaken a genuine fossil for a fake?
I guess it’s down to money a better looking fossil (more complete, more feathers, more interesting) will be worth more. With so much material going through dealers and private collectors there a lot of money to be made.
I don’t know if anyone has mistaken a real fossils for a pure fake, but recently for example a colleague of mine correctly spotted that part of a fossil was fake. But he spotted the wrong bit – it was a different part that was fake than he thought. So he was right for the wrong reason, but had spotted the tell tale signs of tampering which made him check more carefully.
Sometimes fake fossils are done for money, but sometimes they’re done because amateur fossil hunters found a broken fossil and tried to piece it back together.
That’s what happened with Irritator of the Baryonyx family; amateur hunters had found the skull so fractured that they tried to fix it with plaster of Paris. Figuring out what the true shape of the skull was so frustrating for the scientists they contacted that the scientists named it Irritator.
Actually in the description of Irritator in JVP they describe it as having been fixed with, of all things, car body filler. But I wouldn’t call this a fake, just a very bad repair job. The point of a fake is to actively deceive, and I don’t think that’s true of the Irritator specimen. Making something look better than it is or for profit (by extension) seem to be the primary motivation.
Sorry, but saying that the same chimeras can be widely found in other Museums outside China is not acceptable. The fossil collections of most of the Museums around the world are curated with proper scientific criteria, and any visiting palaeontologist can trust and check in the validity of the specimens there stored. We ourselves have visited many such collections over the years and have never come across any fake or forgery. Unfortunately, although in most of the cases these Chinese fakes are recognised with ease by a specialist, they are exhibited in the Museum halls as real specimens, which causes misinformation in the public. That inflicts irreparable damage and loss of information in originally valuable fossil material.
Well they are widely found though. I think you are misrepresenting my position and missing the subtleties of the point.
And be careful with the term ‘widely’ – I’m using it to mean that this happens around the world, not that these are necessarily common. And don’t confuse chimeras with fakes. Go to the Humbolt Museum in Berlin – most of the dinosaurs on display are chimeras – the mounts are made of bones from a number of different individuals of *one* species, but not individuals of different *species*. These are NOT fakes and the material was (largely) well catalogued before it was mounted so we know which bones from which specimens went into this mount. But they *are* chimeras.
You said “We ourselves have visited many such collections over the years and have never come across any fake or forgery.”
And I never said any different. You are conflating a chimera with a fake and the two are not synonymous.
Modern museums are very good for curation no doubt, but this was not always the case (Brontosaurus?!). You can see mounts and specimens like this in many countries where the material dates back to the late 1800s or early 1900s. I have seen or know of material like this in the US, Sweden, Mexico, Peru and others. People made mistakes then and didn’t always catalogue what they did so they remained for a long time, or remain now (even if we know what the error is). So they are indeed widely out there and in museums that are ‘good’ and ‘careful’.
Moreover, i myself say in this post that the fakes and problematic things are in the dogy collections of so-called museums. The good and well run and expertly curated museums in China don’t have a problem. You seem to be accusing me of saying things I didn’t and holding a position which I don’t.
Obviously, my intention was far from misrepresenting your position. I meant exactly what I said, this is, that the abundance of fake fossils within the Chinese collections is a serious problem, with no comparable among the American, Eurasian or African research collections. And please, be careful with a colleague, and don’t misrepresent my position: I know what a chimaera is, and what a fake is. The problem is when some chimaeras mix so many different things that become a fake…
” that the abundance of fake fossils within the Chinese collections is a serious problem”
But in this I think you are quite wrong. The Chinese collections are not full of fakes and I have seen an awful lot of them. There are lots of fakes in CHina, but NOT in properly curatoed museums. In private collections and in self-proclaimed museums yes. But propely run, public funded institutes, I have yet to see anythign that is a fake that is not known to be already (i.e. people knew it was a fake when they got it, but obtained it for the importnat material that was included [like the top half of ‘Archaeoraptor’] or becuase it weas given to them).