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

Re: Coelurosaur Analysis Update featuring Buitreraptor (Theoretical)

I am splitting my reply to Mickey in two parts, theoretical and practical. This
is the former.

Mickey Mortimer (mickey_mortimer111@msn.com) wrote:

<It was merely a hypothetical example of a character which could be subdivided
into multiple ordered states.  I don't even have it in my analysis yet, as I
noted later.>

  My question was about the theory of character dependancy or relatedness. It
wasn't to argue for the influence of the structure on the analysis presented, a
real test of which would require a better analysis with complete data (as
Mickey has noted, this current analysis excludes whole parts of the skeletal
anatomy for the sake of coding complete species as he works on refining the
character studies). Given the use of a single bone in explicit character and
variation studies (which I have been working on for the humerus and pubis for
the last several years) one finds that many characters that are used for states
of shape can be split up since the parts of the whole are more variable than
the expression of the whole in an analysis presumes. This leads to misleading
data in using complexes as useful, and argues for splitting complexes of
features up. 
<Wilson argued this in his thesis, with some sketchy developmental support.  I
actually asked the DML their thoughts on it months ago.  Ordering makes
intuitive sense, as not doing so would let intermediate states be synapomorphic
of clades.  You'd get support for 'Protopteryx + Longipteryx' exclusive of
ornithothoracines, because they both have two phalanges on manual digit III
instead of one phalanx.  Or you'd get support for 'Protoceratops +
Montanaceratops' exclusive of Ceratopsidae because they both have eight sacrals
instead of ten.  If someone would argue this, most paleontologists would claim
the intermediate states are symplesiomorphies compared to the derived ones, and
thus the character would need to be ordered.>

  But unfortunately, the very claim for this assumption of the continuity of a
character transformation is an assumption one has knowledge of the state
acquizitions, and doesn't allow the taxa arrangement to find a more
parsimonious arrangement. In this case, Wilson is correct and Mickey is wrong.
Mickey argues that *Protoceratops*'s eight sacrals would put it outside of
Ceratopsia because other ceratopsians have a different count. But if that
character state was as equal as the others and not a sequential acquisition
without other value, it would be a drop in the bucket with other characters
weighting towards a different explanation. And perhaps it may find a
*Protoceratops* + *Montanaceratops* clade. This does not mean it is less likely
to be true, and the use of this as an example is an argument from incredulity.
Wilson used developmental conditions to suggest that animals may develop the
same count or condition convergently, and that perhaps weight for the condition
is weak. Ankylosaurs, pterosaurs, and some birds even fuse sacrals that never
touch the ilia, forming a synsacrum structure with a "sacrostyle" process of
sacrals which do not possess distally broadened transverse processes or sacral
ribs yet contact the ilia because of the broadening and rotation of the iliac
blades into subhorizontal shelves that thus contact the distal ribs. Is this
the same as lengthening the count of sacrals in ceratopsians which possess
distally broadened ribs contacting the parallel ilia? We know this is
convergent due to phylogeny, but does this not permit us to think the sacral
count homologies are not so useful either? No second guessing, on with the


Jaime A. Headden

"Innocent, unbiased observation is a myth." --- P.B. Medawar (1969)

Yahoo! Mail - PC Magazine Editors' Choice 2005