I wouldn’t mind if the cladists just messed in their sand-pit and left the joined-up-science to the likes of me. But as theirs is a social game, it’s more important for them not to allow their core beliefs to be contradicted than to explore a variety of viewpoints, as you would in real science. This is why they block opposing views if they feel they have a half-reasonable excuse, and they do it wherever they can, not just in the peer review situations we’ll be dealing with at length below. Just as Rhys Morgan was thrown off a medical discussion group for criticising the violations of government health guidelines he found there, almost everyone departing from ‘the norm’ has been thrown off the dinosaur mailing list. I was glad to go, and got myself thrown off when I’d had enough of it but Olshevsky (“Dinogeorge”) was simply removed because they all got fed up with his not accepting the mob’s views. See if you can tell which of his July 2nd postings was so unacceptable here:
Tracy Ford’s name couldn’t be mentioned explicitly in the postings because he’d been thrown off too. (Ford might well have had the biggest good idea ever in dinobird science if it really was him who first suggested the ex-flying status of Velociraptor/Deinonychus. Don’t want insightful scum like that on your science discussion group, now do you.) )
But the worst example of pissing down our backs and telling us it’s raining, is when it happens in review. Of course I wouldn’t waste my time going for a job in palaeontology but there’s no harm in seeing what happens if you try to push some truth through in the form of a scientific paper.
So many good new ideas, perhaps most, disappear into the reviewers’ waste bin with a noise like caterpillers fighting. This is because poor scientists hate interesting new material for the same reason cats hate vacuum cleaners: they don’t even have to think to hate – they instinctively feel threatened.
Just as ordinary book reviews offer a unique window on any area of activity, peer review of papers sent to academic journals is, alongside grant review, the crucial point of impact of decisions directing a field. Where anonymity protects any kind of control system, there is endless and notorious oportunity for pointless damage. Here we see it in operation.
In about 2006 an oportunity seemed to arise for publishing as a paper in a journal, the material I had been preparing to include in a book. I had never expected the faults I was pointing out in the conduct of dinobird palaeontology to be allowed to pass “peer review” by anyone working in the field, as I took a very dim view of their academic standards, and didn’t expect them to accept criticism of those standards or their methods under any circumstance. However, I thought that agreeing to produce a paper might help my productivity, and I would get either a paper published or some incriminating feedback. By 2008 (some tasks need slow care (-: ) I wanted confirmation from the journal on formatting details, so I emailed the unfinished paper as a sample to allow the editor to specify adjustments, mentioning that the version was a long way from complete, needed serious amending never mind polishing, and must not yet be sent for review. I had earlier stressed that any reviewers would have to have a qualification relating to the philosophy of science, and that under no circumstances should it be sent to someone who had gone from school, via a geology degree, straight into palaeontology.
The editor not only did send the unfinished paper for review, but to exactly the sort of reviewer I had expressly vetoed. (By some quirk of motive, in different ways, the editor and the reviewer each made the reviewer’s identification straightforward.)
Here is that review, with my comments…
It is very difficult to understand exactly what points are being made in this MS.
Few people have ever accused me of mincing my words! The paper started with an overall outline of the field, and problems in the way it had been approached in the past. For the more general reader, brief introductions to the types of dinosaurs were given, as the journal was not palaeontological. As in the book, sections on the way theorisation had been approached, and should have been approached, were included. The final sections covered the evolution of feathers and bird-style breathing, and ended with the detailed considerations of “the past” of the dinobird field and how “the future” of it should be improved. As well as offering a new account of the evolution of bird-like dinobirds, and a new phylogeny, (surely highly noticeable! I even drew a diagram!) the middle sections included the results of a Monte Carlo approach to estimating the validity of dinobird cladograms. Though interesting, I had already decided they were inconclusive and was in the process of redesigning the experiments to avoid flaws in the approach (which needless to say the reviewer did not notice).
It seems, after several readings, that the principal novelty of the work reported in the MS consists of a Monte Carlo analysis …
The entirely new phylogeny complete with an entirely new form of phylogenetic diagram – a parlogram – was not noticed by the reviewer. For him, novel explanations don’t count as any kind of science; all the old ones were perfect, so anything new needs new evidence. It was only to be expected that explanations as to how the old mistakes came about should be undervalued if no mistakes were admitted. As part of this kind of review/revamp of the field, analyses of existing theories would actually be essential, but of course they will have been seen as waffle.
Providing a solution to the riddle of the dromaeosaur tail didn’t even count as a novelty apparently. (I’ve seen whole papers written on topics such as this, and seldom making any sense.) The list of half-a-dozen possible flight configurations for “hermeans”, accompanied by comments, apparently did not count, though a number of papers have covered only this topic, and less exhaustively. Did the reviewer think some kind of computation is required for anything to be scientific? Probably, when it suited him; ironically he himsef is not very ‘computational’.
Of course any of my arguments considering functionality were certainly not counted as science, particularly when used for ellucidating phylogeny, even though possible steps in the origin of genetic mechanisms (the evolution of altruism and sex, to name but two) have been checked for plausible fitness for decades. The relative impossibility of evolving a blade-shaped feather from a “down” type format, compared to a flat blade, did not count as a possibly useful novel insight, nor did the sigificance of bipedality and warm-bloodedness in the evolution of feathers.
…carried out on two cladistic datasets, one of which is published and one of which is not (the author of the unpublished dataset is not acknowledged, so it is not clear if the author of the MS had permission to carry out the analysis and publish the results).
A lie. The authorship of each cladogram was acknowledged. And I don’t know why the reviewer wanted to suggest permission would be needed to input data from or details of an experiment into a new experiment or even report! It is because replication and suchlike are part of the routine of science that details are given as fully as they are. As for publication, whatever the special definition given by bodies such as the ICZN (which governs the naming of new species) for their own purposes, if a work can be obtained through standard Inter-Library Loan, then it’s been published.
However, I could not determine from the MS the significance of the results of the Monte Carlo test.
Unlike with cladograms, my results actually reported statistical significances! But though the paper was unfinished, and as I stressed, was not yet ready for review, I had made the meaning of the experiments clear as I understood them at the time… though I wouldn’t expect the reviewer to understand them.
Presumably figures 1 and 2 are graphical representations of the results, but what do they mean and how do they compare with the original cladograms?
The paper had not even been finished, but in fact the explanations were pretty good, especially for people with any kind of hinterland in simulation. Even if justified though, this criticism would not be grounds for rejecting the paper.
Much of the rest of the MS consisted of unreferenced or poorly referenced…
By that time the paper had 202 lines of references – about five pages, compared to 26 pages of text, and 12 pages of graphics and tables. I have no doubt most readers will feel the work is far too cluttered with references. Paul’s “Dinosaurs of the Air” has 26 pages of references, after 410 pages, about half of which is text.
…subjective assertions often lacking any supporting evidence or data…
The offering of new explanations is always dismissed in such terms by those for whom science is either the output of a cladogenesis program or the drawing and descripition of a fossil, rather than its true nature: the production of progressively improved explanations. My definition of evidence is not just more detailed than the reviewer’s (see
sciencepolice-14) but more cogent.
…and riddled with inaccuracies, or in some cases just plain wrong.
I’m just going to say “rubbish” to that – unless an example is provided. Ah – here’s one. How good is it, I wonder?…
Take for example the statement regarding oviraptors:
“They were originally thought to have eaten eggs, and some forms probably did”.
The discovery in Mongolia in the 1990s of an adult oviraptor preserved in situ on top of a nest with eggs, followed by several additional, almost identical associations (e.g. Deeming, 2006, Palaeontology 49: 171-185, and refs therein), led to the now widely accepted suggestion that these oviraptors were fossilised together with their own nests and were not stealing and eating the eggs of other dinosaurs, as had long been thought. This idea has been very widely broadcast consequently the authors apparent ignorance of this is incomprehensible.
Not just a staggeringly facile criticism – blatantly invalid, as any half-way reasonable interpretation of my words will confirm. What did the reviewer think the word “originally” was for? Actually, the little complication of “first we thought they did, then we thought they didn’t, and now we think they may have done after all” is hardly essential (though I did hint at it). I wonder if the reviewer thinks their diet didn’t include eggs? He’d be in a minority, and rightly so. If that pathetic fabrication of an example is the best he can do, we needn’t take any of his accusations of inaccuracy and error seriously.
The MS also suggests that the author has an incomplete understanding of both modern phylogenetic systematics…
My identifying and detailing of undeniable dangers in phylogentics that are understood and discussed by genuine experts but routinely ignored by dinobird palaeontologists, does not constitute an “incomplete understanding” of modern phylogenetic systematics. Emphasising the widespread ignoring of these dangers doesn’t either. It would be odd indeed if someone like myself with a history of involvement in complex information processing including academic research and an MSc in the subject (as well as reading books on cladogenesis) had less understanding of the principles than someone like the researcher who started using cladograms with just the background of a geology degree. It’s not surprising that he fails to include a single technical, scientific or philosophical principle in his critique (which actually can hardly be described as any kind of academic critique). Writing my own cladogenesis system from scratch using two different algorithms I’d devised myself meant that I had to have some understanding of it afterwards if I hadn’t beforehand (though of course I had). Of course, running experiments explicitly demonstrating the ways homplasy distorts the results would have given my understanding more detail but I didn’t need to run them as not only did I already appreciate the principles, which had already been demonstrated over and over again by others, but even that demonstration would again have been completely ignored by the reviewer and like-minded cohorts.
In what he writes next, the reviewer is actually making the ‘argument’ that cladistics is widely used; the implication apparently being that if it’s used it must be ok. It was because I anticipated this kind of meaningless waffle that I stressed the importance of having a reviewer with some philosophical component, rather than merely a propensity to express “how dare you” as pompously as possible.
…and of palaeontological research. During the last three decades phylogenetic systematics has grown into a very powerful set of ideas and methodologies widely used throughout biological sciences, and increasingly in other disciplines such as, for example, linguistics and vulcanology. There is a vast literature on this discipline…
A fair amount of good work is indeed being done into the theoretical background of phylogenetics – which is why I mentioned it (along with a few such workers) in the paper, and as I said, none of it ever seems to reach dinobird palaeontology. Does he exemplify any dinobird cladogram (not counting modern birds) published say in Science or Nature in the last fifteen years, that included any acknowledgement at all of the dangers of homoplasy (convergence/parallelism) ? He couldn’t, even if he wanted to. Any use of cladistics falls into three categories: where distorting influences are understood and taken account of; where they aren’t understood but are absent, producing reasonable phylogenies; and finally where they are present but not taken account of, resulting in false phylogenies. But of course, with all phylogenetics, it is rare that any error is discovered, and far less that it’s costly enough to matter. An exception is in epidemiology, where the progress of disease needs to be well understood, and my paper included an example where a virus had been tracked and its phylogeny found to violate parsimony. Because feedback on phylogenesis is so rare, many “ideas and methodologies” may indeed be developed, as they are in astrology, with no guarantee of validity; indeed pressing on while ignoring the best ideas guarantees the invalidity we so often find in dinobird morphological phylogentics.
…and issues such as the rooting of trees have been discussed in great detail by many many authors (see for example reams of paper in Cladistics, Systematic Biology and other journals). The account given in the MS of the problems of tree rooting does not in any way reflect current understanding of this aspect of cladistics and the author should consult the relevant literature before venturing into print on this topic.
All that “current understanding” stopped those hugely flawed bird and mammal morphological trees being published did it? No it didn’t. I do hope he and his colleagues haven’t spent much time consulting the literature since if they have it’s been wasted! Not only do they ignore new developments and endlessly repeat obvious errors in this area but they never even seem to employ its jargon beyond that mentioned in the tabs and options of their cladogenesis programs. If I had £1 for every dinobird palaeontologist who behaves like they understood that high bootstrap support does not guarantee freedom from homoplasy, it still wouldn’t pay for a train ride. There is no use of an outgroup to root a subtree that does not involve and indulge prior assumptions about the connection points of both subtree and supertree. Have dinobird palaeontologist stopped using outgroup rooting yet? Oh, and it’s “…consult the relevant literature” is it? I’ve read most of Felsenstein 2004. Has this guy? Don’t make me laugh; he’s probably never even seen it.
A striking feature of the MS is the extent to which the author derides palaeontologists and palaeontology…
This is why it is essential to start with that issue: Give one reason why someone is wrong, and people will shrug. Give three reasons and they tend to be convinced. But give a hundred ways that hundreds have been wrong for decades, and people will wonder how so many could really be so wrong for so long. Not that valid examples couldn’t be given where it was true, for any number of people, ways and decades. Given that the number of TV and radio channels dedicated to religion and ‘the unexplained’ exceed those devoted to science by two orders of magnitude, and the number of creationists in the world vastly exceed the number who understand evolution, it should be clear to most scientists that simplistic number comparison doesn’t decide scientific issues well. Also, most significant new theories effectively project all past practioners into error. Nonetheless it was still vital to explain that all the errors I mention boil down to very few basic reasons: the difficulty of checking theories in palaeontology, the wide spread of subjects it relies upon, and the low level of specialist expertise amongst the average palaeontologist in the vital basic subjects. You can also add low moral integrity to that.
…and yet large swathes of text are devoted to wild speculations on the functional morphology and ecology of dinosaurs and birds of the kind that palaeontolgists have, to a large extent, successfully eschewed in recent years. See, for example, the review by Hutchinson and Gatesy (2006) in Nature 440: 292-294.
(“Successfully”?! Yeah right! By what criterion?) Does he mean the bit where Hutchinson and Gatesy say: “The subject of dinosaur locomotion requires palaeontologists to become better functional biologists — experts in more than just skeletal morphology”? Does that support the view that functionality-based arguments are necessarily wild speculations? Try proving that principle. (As a general rule, expect any ‘scientific’ criticism that relies on accusations of speculation, to spring from a failure to understand that “conjecture”, i.e. speculation, is half of science. The best science comes from the use of intelligence, which is itself built from good use of unprovable heuristics.) Maybe the reviewer accepts functionality can be used for some science but not phylogenesis. As I mention above, better scientists than the reviewer have taken the evolutionary significance of functionality seriously for decades. I don’t always put figures to my theories but you don’t have to finish a project the moment you start it. When I say for example that a dromaeosaur’s tail could be used to accumulate destructive energy, it’s immediately clear to most people with a reasonable grasp of physics that it might well be true or is at least worthy of further investigation, whether I’ve conjured up some set of figures or not. I might add that I’ve seen a great many numbers used in vain in dinobird science.
Hutchinson and Gatesy say: “…it is essential that the questions do not get too far ahead of the evidence, that tides of enthusiasm do not drown concerns about ambiguity, and that novel tools are not recklessly applied.”
Cladistics has been recklessly applied. Don’t take my word for it, ask even Fortey, or of course Sober. Readers should though be warned that this is the same Gatesy whose disgraceful panning of Paul’s “Predatory Dinosaurs of the World” shot me back into palaeontology, and whose approach I’ve already had to correct in my book. He seems here to be warning people not to come up with too many theories, perhaps until the likes of Gatesy have verified those that have gone before, in case everyone gets enthused with the wrong idea. (People can actually come up with any idea, though it may of course be refuted, and those whose ideas are frequently genuinely refuted will be progressively less listened to by the worthwhile scientists.) Unfortunately dinobird palaeontology is nothing if not made up of unjustified ideas enthusiastically propagated by those who should know better but never seem to. Hutchinson and Gatesy seem slightly embarrassed to be writing a paper involving functionality, having scorned its usefulness in the past, so they add a little equivocation. Not for the first time we see the unhelpful use of the words rigorous and testing. It is this air of disapproval the reviewer is referring to. But we’re not dealing with a minor irrelevance here. The kind of peer review I’m having to deal with, and the kind of review Gatesy co-authored of Paul’s work, are the background to Paul currently having his contributions to the Journal of Vertebrate Paleontology persitently rejected, and according to some, references to his work routinely removed from books. We see the science move into a third decade of having its best ideas, and in at least Paul’s case its brightest stars, dragged down by talentless pseudo-scientific scum.
As to whether use of “functional morphology and ecology” have been “successfully eschewed”, if Hutchinson and Gatesy attempted it in the example given it would have been hypocrisy; more importantly, convincing any number of already convinced incompetents is not the mark of a successful scientific argument.
Finally, the author’s case is not helped by numerous deficiencies in the text including misspelling, omissions of captions to the figures, and a very poor standard of syntax that often leaves the reader at a complete loss as to the intended meaning. What are we to make of the following example:
“Even now there is still no good example of an earlier feathered type capable of powered flight, and although none the specimens of that type that we have will have been the first individual capable of flight, and indeed Archaeopteryx may not have been the first genus with powered feathered flight, it is still reasonable to believe the urvogel (the first powered flying bird) was closely related and not many millions of years earlier.”
Oh dear! It’s a good job Proust and Julius Caesar didn’t have to get their sentences past this reviewer! It would have been more helpful if he’d pointed out that “of” was missing after the “none”. It might come as some surprise to him that people in the writing profession repeatedly approve of my work. I was winning prizes in essay competitions before the reviewer was born, and he’s onto a loser if he claims I shouldn’t be published because of the hopelessness of my prose. An expert speller took a look over the text immediately afterward and found 12 typos – not bad for 26 pages prior to any kind of spell checking.
Or in a section commenting on therizinosaurs:
“Often misunderstood, they may have evolved at one or several different places, perhaps not from very similar ancestors.”
That’s a serious error? It originally said “… evolved at one or several different times or places” so the “at” might have been replaced by “in”. Some writers benefit from good editors, and that kind of correction is part of their job, but its usefulness doesn’t make any writer unpublishable. As for typos and captions, if the paper had been ready I wouldn’t have told the editor not to send it for review.
Undoubtedly there is plenty of scope for deconstructing the history of research on the debate over the origin of birds, and for analysing the philosophies, methodologies and personalities involved.
Or you could say the science has been dead for decades and needs to be rebuilt from the ground up – without involving the likes of the reviewer.
Moreover, the utilisation of alternative techniques such as Monte Carlo simulations for assessing the results of phylogenetic analyses is also to be encouraged.
Apart from the ridiculous situation of judgements and advice offered on this topic by someone with the reviewer’s qualifications, there are important issues here. Research tools need to be calibrated; for example an airborne radar technique for estimating polar ice thickness would need to be compared with readings from other methods before its outputs could be interpreted properly. This has never really been done with cladograms used for dinobirds, and it can’t be done without simulated evolution under various conditions, and cladogenesis on the results. Never mind “alternative techniques” and “to be encouraged” – it should be clear to any good phylogenists in the area that it’s essential, and they should have been spending at least half their time investigating simulation methods for identifying homoplastic distortion of cladograms and its removal, until a solution were found. I don’t know any dinobird palaeontologist who has ever even mentioned doing this, far less attempted it, and just that on its own is enough to damn the whole field.
However, the MS at hand does not make any meaningful contributions to either of these endeavours, is far more likely to mislead than to inform and, as it stands, is wholly unsuitable for publication.
You can hear the sanctimonius slime slithering off every word. Not one valid example given of the MS misleading, and half a dozen meaningful contributions not recognised, each of which on its own more than sufficient to justify publication. Nonetheless, what the review does reveal is the dire mentality underlying current dinobird orthodoxy, and even more vividly portrayed is the menace of peer review itself. The reviewer is most certainly not the peer of any significant transformer of the field; indeed he has shown himself unable even to recognise any of the worthwhile innovation I offered.
Subsequently I sent the paper to a journal with a more open-minded reputation, but from the short list of reviewers offered and the reviewer’s summary which, though brief, made his identity tempting to guess, the reviewer was from exactly the same category as the first. The second one did though take the trouble to say my philosophy of science was rubbish. Ironically, that isn’t the style used by those genuinely into philosophy: they’d have been unable to resist sub-defining “rubbish”, and then offering a succinct demonstration of at least one example of it in the paper.
Of course, I couldn’t expect that which I call a turkey to vote for my Christmas, but even if it were to affect the reviewer, nothing can justify blocking progress if the blockage wrings the neck of a science year after year.
That review demonstrates the absolute uselessness of peer review to a science in a state of absolute putrescence, as well as the need for the likes of the reviewer to be shunted into a siding of their own, where they can’t interfere with those of us doing joined-up-science more in tune with the scientific community at large.