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

RE: FW: Dracorex's phylogenetic position examined with science

Denver Fowler wrote-

I think it is pretty fair to say that the "examined
with science" tag to the email title should have read
"examined with cladistics". The particular choice of
wording of the email title could be read as a personal
attack on Bakker & Sullivan and their credibility as
'scientists'. in this case it has certainly caused
offence, whether intentionally or otherwise.

Honestly, I don't think Bakker et al. or Sullivan (2006) perform phylogenetic work very scientifically in these papers. They call previous results into question for illogical and non-sequiturish reasons. Bakker et al. then explain their phylogenetic scenario, which uses very few characters, all of which are in published pachycephalosaur matrices anyway. They defend this with a lot of highly speculative, perhaps forever untestable, reasoning. It's a scenario sure, but is it a scientific hypothesis? What data could we hope to use to test if "sexual selection acted on the genes expressing the dome"?

There are methods of testing phylogenetic hypotheses
other than cladistic 'objective' character analysis,
and it should be remembered that these can directly
complement or add to said cladistic analyses (eg.
non-cladistic description highlighting new character
states/ innaccuracies in current ones / stratigraphic
occurrence etc). There have been more than a couple of
recent cases where papers have been accompanied by
innappropriate/innacurate cladistic analyses, and this
was roundly criticized, yet sadly this practice
continues. In this particular case a cladogram was not
provided with a published article, as the authors felt
the data was not robust enough, and they were
ridiculed for this.

I never insulted Sullivan or Bakker et al. for lacking cladistic analyses in their papers. I agree not all papers need one, and there are other ways to further phylogenetic knowledge. But Bakker et al. claimed Dracorex created a problem for pachycephalosaur cladistics, saying it made "all previous phylogenetic analyses ... inadequate." This is quite a strong claim to make without testing.

I don't necessarily see myself as an anti-cladistic
person, but there are times when the significance of
such work is overstated (especially when contrasted
against specimen collection and analysis). I'm not
about to perform a phylogenetic analysis (i am too
busy out in the field helping collect new specimens
for analysis), but surely the most scientific way to
refute Sullivan/Bakker's claim would have been to use
their own dataset to prove them wrong? It's pretty
clear that Sullivan & Bakker don't consider Williamson
& Carr's character matrix as correct.

Well, Bakker lacks a pachycephalosaur matrix (as I'm sure he always will). Sullivan's (2003) is based on Sereno's, with several characters added. From prior experience, I know Sereno's analyses tend to exclude conflicting data, giving them very high Consistancy Indices (Sullivan's variant of Sereno's is .91 compared to Williamson and Carr's .83). So not only did I feel Williamson and Carr's would be less biased, but I felt it would have the greatest chance of changing to a novel phylogeny (as advocated by Sullivan) given the data added by Dracorex. It would be interesting to code Dracorex in Sullivan's matrix, just as it would be interesting to merge the two matrices.

Mickey Mortimer