[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]


 today I found these two Cuban pteros skeletal drawings:


 The Nesodactylus skeletal drawing is pretty good. Anyone of you know
 the author of these drawings?

The author isn't mentioned on those pages, but then I saw the side bars. They assign lots of Paleozoic tetrapods to Archosauromorpha or Lepidosauromorpha.

Then it dawned on me:

This is David Peters' site.

When his manuscript on his analysis of amniote phylogeny was rejected, he didn't do what normal people do -- try to get the problems out of the manuscript and the analysis, and resubmit --; instead, he put his stuff on the web.

His name doesn't seem to be mentioned anywhere except at the bottom of this page: http://www.reptileevolution.com/about.htm

Indeed, the top of that same page claims: "_This website is a big colaboration [sic]._ This site would not have been possible without the hundreds of paleontologists, preparators, paleoartists and editors who discovered, exposed, reconstructed and published the taxa presented here." This is utterly misleading; of course Peters stands on the shoulders of giants, but to present them as his coauthors, as if that website were a collaboration project by the entire field and thus presented the broadest possible consensus, is just not honest.

Also on that page, Peters says:

"_Sometimes I'm wrong._ Sometimes others are wrong. Most of the discoveries herein were made simply by expanding the taxon list. In other cases purported autapomorphies turned out to be misinterpretations (exemplified by my published mistakes with Cosesaurus [link] and Sharovipteryx [link]). If you find misinterpretations here, or new data becomes available, please contact me [link: info@reptileevolution.com ]. It's more important to get it right than to be right. Credit will be given!"

That's nice, but completely impracticable, because to seriously dig into the problems in Peter's tetrapod analysis would be a MSc thesis or more. Peters relies on his photointerpretation technique (read: naked, shameless pareidolia), does not even try to avoid correlated characters, and, well, there are "210 taxa and 228 characters" -- that's _way too few characters_ for so many taxa. I'm currently working on a matrix with 102 or 111 taxa (I added 9) and 289 parsimony-informative characters (down from 333, because so many were correlated), and even for this much smaller number of taxa I am reminded all the time that lots of important-looking characters are not in the matrix...

Maybe I'll just mention that Kat Pawley's (2006) thesis claims that *Casineria*, which was described as a close relative of Amniota, is indistinguishable from *Caerorhachis*, which could be the sister-group to all other temnospondyls or even outside the temnospondyl-seymouriamorph-diadectomorph-amniote-lepospondyl clade. In the matrix I'm working on (see my thesis, 2010), they are indeed indistinguishable. Not having seen the fossils, I'm not going to formally synonymize them (never mind that I don't want to scoop Pawley either), but somebody should have a long, hard look at them.

When I treat *Casineria* and *Caerorhachis* as the same taxon, it comes out where I just said *Caerorhachis* alone does.