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

a frustrated phylogeneticist's complaint

Ken (and others),

As there is no way to assess the "quality" of a character used in phylogeny reconstruction, I would have to argue that each and every one of the characters used by Sereno (as well as those used by other authors) are equally informative. "Strong" and "weak" simply have no meaning in this context - either they are informative at one level, or not (and informative at some other level).

We can discuss the robustness of a node, but not of a character. And if the characters are homoplastic - so what? If repeated analyses of character-state hierarchies repeatedly get the same result, AND if the list of characters renders it robust to standard measures of nodal support (Bremer index, bootstrapping (and other permutation tests), jackknifing, etc.), then the node is robust. Not fixed, but robust - all nodes are subject to rejection if we look at new character systems or taxa. At this point in time, nothing we've seen leads us to suspect that Dinosauria's content is in any danger of being overturned, and we can regard Dinosauria as a stable clade. Maybe someone will discover that turtles are actually derived ankylosaurids, but until then (i.e. in the absence of a competing hypothesis, as several others have pointed out), it is meaningless to question Dinosauria's current content.

Your point about characters that, as defined, lack clarity has merit. One problem is that Sereno's matrices at this level have, thus far, been published in media that do not lend themselves to thorough discussion (presumably with clear figures). But presumably, one can look at the taxa Paul looked at and get some sense of what he was discussing (though not always, which is why I agree with you on this point). One could also contact Paul directly and ask him what he meant - people have done that to me on occasion, as I've been guilty of unclear discussion. (and by the way, nontherapsid synapsids do not have posttemporal fenestrae, as far as I know.)

Also - from this discussion, one might conclude that Ornithodira is overturned. Not even close, folks.


Anybody can take two taxa and use them to define a holophyletic clade. The question is whether they should. Clades should be named for groups that appear to have strong synapomorphies, and to me "postfrontals absent" cannot be defended as strong. A long list of weak characters isn't convincing to me, especially when reversals are automatically invoked whenever the weak characters are not congruent.
Question. If the three dinosaur "synapomorphies" not found in sauropods aren't reversals, and they lie outside the Triceratops-bird defined "Dinosauria", do we then "educate" traditionalists and the public that sauropods are not dinosaurs, or do we redefine Dinosauria?
And I still haven't gotten any good answers to which of Sereno's dinosaur synapomorphies are the strongest (clearly separate dinosaurs from non-dinosaurs). Just two or three (is the long deltopectoral crest pretty reliable?) I've already criticized "postfrontals absent".
What about posttemporal opening reduced to small foramen? Compared to what? How small is small? Are there no "small" ones in any basal crurotarsans? Someone posted that this sounds like a strong synapomorphy, but admitted not even knowing what it is. Was that supposed to be convincing?? Is there anyone who knows more about it who wants to defend this synapomorphy? Aren't pterosaur and pelycosaur posttemporal openings small? I see no reason to trust such a character.

Christopher A. Brochu
Assistant Professor
Department of Geoscience
University of Iowa
Iowa City, IA 52242

319-353-1808 phone
319-335-1821 fax