Science is fascinating and fun. People enjoy doing it and many are those who feel called to explain it.
Too many. Science requires not just concentration but wisdom and a frighteningly open mind, because it’s a deceiver. The unwary and wary alike will encounter pitfalls, and science is all about escaping and avoiding them. It is irresponsible though to guide others into errors, particularly when they’ve been pointed out to you time and time again.
Take Carl Zimmer for example, who often writes about palaeontology. Whatever part of palaeontology you’re in, you’re always a million miles from a proof. And the further a science is from a convincing experiment, the more vital it becomes to theorise responsibly. It’s also more vital to watch out for the effects of human corruption: just as people with all their needs supplied tend to engage in social games and competition (often disguised), a science trying to guess at arbitrary past events can become dominated by social influence.
How bad can it get? This month in the National Geographic, Zimmer propagates a blunder he can’t have spent any time considering. He relates how tree-living animals might have become gliders as their scales lengthened into feathers, which were then perfected further for powered flight. He’s comparing that against the idea that running animals learned to fly by simply flapping their arms. Each explains the similarity between birds and bird-like dinosaurs: with the gliding (“trees-down”) theory, birds may have lost the power of flight and produced dinosaurs with bird-like features, including feathers. With the “ground-up” theory, non-flying ground dinosaurs were the source of birds.
Notice how any similarity between bird and dinosaur is explained and predicted by both theories, yet in one case bird-like dinosaurs came from birds, while in the other, the reverse happened.
In science, we delight in making observations that help us decide between two theories. Litmus paper for example turns pink in an acid and blue in a base. Each result is consistent with just one theory. How clumsy do you have to be to base your choice of theory on a test where the observation is predicted by both theories? As clumsy as this:- Zimmer says, of the “trees-down” theory: “This feathers-led-to-flight notion began to unravel in the 1970s, when Yale University paleontologist John Ostrom noted striking similarities between the skeletons of birds and terrestrial dinosaurs called theropods…”
What actually unravelled then was our unquestioning respect for Ostrom’s powers of deduction, and that of his co-theorists, since Ostrom then (like Zimmer now) inferred that mere similarity between two animals says something about direction of descent. The folly is highlighted by the ground dinosaur he was studying, Deinonychus, living 80 million years after Archaeopteryx, the first bird… yet to Ostrom at the time, and Zimmer now, it is “obvious” the early bird was descended from the later “dinosaur”. It’s no good saying Deinonychus’ ancestors 85 million years earlier might have led to Archaeopteryx – yes, that is conceivable, but the point is that Ostrom completely ignored the more straightforward explanation, and Zimmer wants to force us to keep making the same mistake 40 years later.
What actually happened was that originally there were only two theories: either birds came from dinosaurs, or they are not closely related at all. Mere similarity would then have been a significant test; the presence of aberrant feathers on ground forms like Sinosauropteryx would have meant feathers possibly hadn’t evolved for flight. But in the mid 1980’s Tracy Ford and Greg Paul[2, 3] realised there was a third possibility: some dinosaurs may be flightless birds.
(Let’s pause for a moment here to appreciate how incredibly unskilled and unprofessional it had been for that possibility not to have occurred to palaeontologists at Yale, the natural history museums and a hundred other institutions, for at least ten years. Paid to apply expert science to the problem, all they had to do was either to write the first theories down as diagrams and notice the third that was part of the full set, or draw conclusions from the loss of flight in modern animals, but they couldn’t manage either.)
Soon afterwards, George Olshevsky realised it wasn’t just bird-like dinosaurs that might have come from birds: all dinosaurs might have come from much more primitive gliding ancestral birds. I too thought that on reading Paul (1988), wavered a little, then realised it was right. So now there were two ways in which birds might not be descended from dinosaurs, but in the new way, similar and intermediate forms between birds and dinosaurs are predicted.
You might ask yourself why Zimmer doesn’t bother to tell his readers of the ‘new’ development, more than two decades after the fact, but instead just fobs them off with a lie. Why does he violate first one scientific principle – that a test needs to distinguish between two theories, and then another – that all reasonable versions of a theory must be dealt with to refute it? (…as well as the journalistic principle that news must be reported in the first ten years!)
The reason is this. To “cover themselves”, reporters of all kinds traditionally go back to ‘experts’ for their opinions. Or at least that usually happens in science, which is considered beyond the powers of normal people and journalists to comprehend, though in sports, economics, the arts, politics, or even in some scientific branches such as medicine, journalists sometimes think for themselves. In fact, although the tools some scientists use are complex, the basic theorisation is often, indeed usually, simpler than stuff on the money programmes. And frankly, as we saw a couple of paragraphs back, professional dinobird palaeontologists aren’t the sharpest knives in the box anyway. But the palaeontologists keep feeding him the crooked line, and he keeps passing it on to the public.
In fact the true situation is even worse. Deducing how animals evolved by checking all possible family trees and choosing the one with the smallest number of evolutionary changes, sometimes works. It may be surprising that it ever works, but there are good reasons why under special circumstances it can. It’s not surprising though that it often doesn’t; we’ve had many examples where the tree was very wrong. But here again, with dinobirds, we have a test that doesn’t distinguish between theories: whichever theory were true, the computer would produce the tree that it does. Whether or not it’s mentioned, these unreliable family trees are assumed to be 100% reliable, and their structures taken at face value by those Zimmer takes seriously.
Over the last dozen years or so I’ve watched how the errors forged and propagated by these supposed dinobird authorities fall into recurring patterns. The guidelines for the pursuit of the basic principle of science (choose theories that best explain or predict the observations) that they violate can be put into a number of arrangements but I’ve condensed them into 14 basic sections. The idea that the bird/dinosaur similarity refutes “tree-down”, itself violates section 1, sentence 1 : “The business of science is the production, selection and honing of theories which model (i.e. explain or predict) the best” (and others), as does the error in the last paragraph. And thinking that by refuting one version of a theory you’ve refuted it all? That would be violation 6:1 : “A theory is refuted only when all its reasonable instantiations are refuted, not just one”.
But Zimmer’s account produces folly after folly through to the end. For his next trick he presents the evolution of feathers from flat scale, to insulative down(!), and back again towards a flatter form perfected for flight, as “…piecing together a more detailed history of the feather.” If – if – that theory is consistent with some evidence, an alternative sequence from flat fibreless feather format for gliding, progressing towards modern fibrous forms, is fully consistent with all the evidence. Does he mention that to his readers? Does he even take it into account? Does he heck. He therefore picks up violation 10:3 : “ ‘Positive evidence’ must not be trumpeted as ‘evidence for’ a theory when competing theories explain the same evidence.”
He then repeats the mistake he made in choosing one theory of a sequence of feather evolution over another that explained more, by choosing Prum and Brush’s account of the evolution of the cellular machinery for growing feathers. In their view, the amazing system that’s capable of producing in folded up form, an object that blooms into a broad plate, but grown from a base no bigger than the central shaft, was, he says, set on its way by the throwing of a genetic switch. He does hint there was more to it than that, but it hardly needs to be said that alternative (and superior) explanations are available since any evolutionary step could be described as throwing a genetic switch. However, presenting the sequence Prum and Brush actually present in their paper as a preferred theory, is a repeat of violation 10:3. As I recall, before I found out how birds actually did it, I couldn’t even imagine a system that could pull the necessary geometrical trickery, but as it happens, the system we see today is what you’d need for producing an initial form of flat non-fibrous feather (some indeed sported by living birds).
“There’s an even more astonishing possibility” Zimmer says. The possibility that since crocodiles make their scales using a feather-making gene, suggests they once had feathers. I don’t know where he got that idea from – he certainly didn’t get it from me since he makes a point of never listening to a word I say. As it happens, I rather think I was the first person ever to suggest this, but a recurring theme of palaeontological pseudoscience is to revile those with the good ideas, and try to work things round over the years till you can claim you thought of them first. When you do, airbrush the true originator but insult him further by claiming some new piece of evidence as the first genuinely scientific justification for the theory. For the record, the pattern of scales on a crocs’ back matches the pattern of the first feathers on the back of a humming bird chick, which matches the arrangement of plates on the back of Stegosaurus, which matches a possible arrangement of flat non-fibrous glide-worthy plumes on the back of Longisquama. (That’s one of several Triassic fossils they love to hate, that they wish would go away, but which our theory would have to invent if they hadn’t already existed.)
So if the first feathers were insulatory plumes, how does Zimmer get out of the bind that cold-blooded croc ancestors who can’t ever have used insulation, had feathers? Well, he just pretends the problem doesn’t exist. With one bound he was free: “If feathers did not evolve first for flight…” he says, and takes it from there. Completely ignoring their use in gliding because of the similarity he’d mentioned between Archaeopteryx and Deinonychus (!), he astonishingly ignores the possibility that an excellent candidate for a true ancestor to Archaeopteryx – recently discovered Anchiornis – was a glider. Its feathers were for display, we are assured. We are asked to believe that since the time crocs were evolving 70-odd million years before, display was so important that Anchiornis was covered in them, yet at no time was their potential for gliding exploited despite even having masses of long feathers on its legs! It went straight from using them for display, to using them for the much more demanding powered flight, yet without gliding ever becoming evolutionarily significant. (Let me guess… the powerful and complex wing structure and musculature required… evolved by a ‘genetic switch’?) Zimmer assures us that the coloured patterning we now know Anchiornis had on its feathers, is “a big boost” to the idea that feathers evolved for display. Violation 10:3 again, since colouring on feathers is explained at least as well by gliding being their original use. We needn’t add, though we could, that by his stye of ‘thinking’, if he’s happy that crocs had feathers which clearly evolved before Anchiornis, the fact that crocs don’t use colourful displays is reason to believe this wasn’t their earliest use.
How on earth did he ever allow himself to get into this mess? How can he live with the realisation that by a hundred years from now and forever after, people will see those who argue like him as idiots of the first degree? Because for him, only professors count, even if a six-year-old can see the faults in their arguments – even if the issues lie outside areas of expertise of ‘the special ones’ on his list. If it’s to do with ancient birds, even if it’s a question of physics, as with flight, or information engineering, as with judging the accuracy of computer generated trees, not only is a palaeontologist the only person to ask, but others who do have relevant qualifications must always be ignored. (Most paid dinobird palaeontologists qualified only in geology at first degree level. They typically recycle their tribe’s often erroneous traditions in graduate degrees in palaeontology. Zimmer of course does not have a science degree.)
Like many disciplines, palaeontology operates by “peer review” – a kind of “droit de seigneur” whereby members of a self-selected closed shop block the publication of papers that challenge them. It doesn’t destroy experimental sciences because clear experiments tend to cut through waffle. But in palaeontology well-placed figures violate whatever principles of science they like, partly because they don’t know them, partly because they wouldn’t care if they did, but mostly because no lives or money are lost if the dinobird family tree is wrong. In palaeontology, the need to rely on personal judgement has led to the flourishing of pseudo-scientific robber-barons.
Zimmer positions himself between these charlatans and the public, passing all their rubbish unselectively and unquestioningly, and benefiting in the process. He’s established cosy relationships with the likes of Yale, Nature, and of course National Geographic, who are happy to remain associated with his scientific blunders, and who never ask themselves how someone with no science degree is capable of judging the scientific validity of those qualified in multiple sciences. But this has involved him in the death not just of dinobird science as represented by him to the public, but indeed of the whole of vertebrate palaeontology, since the same errors are repeated with his help in whatever sub-field you care to examine. We now know humans walked upright before we split with chimps, and we’ve known this was the best bet for years. Zimmer however propagated the wrong view on that as with most disputes in human evolution, continuously ignoring the views of those visionaries who turned out to be the most accurate.
Interestingly, his area of special interest is parasitism. As when… “a person habitually relies on or exploits others and gives nothing in return”. Yes, he gains his own sustenance at the expense of science, while living within its infrastructure, and specialises in the study of the parasite! At last! Something that makes sense! You couldn’t make it up.
 Paul, G.S. (1984a) The archosaurs, a phylogenetic study. In Third symposium on Mesozoic Terrestrial Ecosystems, ed. W.E. Reif and F. Westphal 175-180. Tübingen.
 Paul, G.S. (1988) Predatory Dinosaurs of the World: a complete illustrated guide. Simon & Schuster. NY.
 Prum, R. and Brush A.H. (2002) The evolutionary origin and diversification of feathers. The Quarterly Review of Biology, 77, 261-295.