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

Re: pterosaurian monophyly



In a message dated 95-09-20 16:31:38 EDT, rowe@lepomis.psych.upenn.edu
(Mickey Rowe) writes (quoting Kevin Padian):

>Since I started this, I figured I should continue it; I forwarded
>George's message to Kevin, and he responded:
>
>  Mickey -- George is of course correct, and his point is frequently
>  raised in systematic discussions.  A cladogram is a starting point
>  for further analysis.  It is a simple tool.  It shows a pattern, but
>  we know that biological patterns are the result of processes.  A
>  pattern is the first step at getting at what processes we might
>  usefully examine next.  There are indeed some instances in which 32
>  characters are placed against 50 in rival phylogenies, though to
>  date these have mostly been molecular analyses.  In more typical
>  analyses, if there is this much homoplasy, then even process
>  analysis may not untangle everything.  On the other hand, if you
>  don't use an explicit methodology (whether cladistic or something
>  that will ultimately replace it), it makes it even more difficult
>  for others to evaluate your results.  (It's like the old saw: if you
>  think education is expensive, try ignorance.)
>  
>  Obviously George is not arguing against methodology, but only
>  pointing out its limits and asking, what next? 

Yay! Right on!

> Analysis of competing phylogenies can be done in several ways.  You can see
>  where the character conflicts are -- what data sets are particularly
>  prone to conflicting signals?  If molecular, do other molecules give
>  similar or different results?

And, by the way--which counts more? Morphological or molecular phylogenies
(assuming they're in conflict, which they often are)?

>  If morphological, are these dental,  cranial, or postcranial?  Do they
involve certain
>  functional or adaptive complexes that might result in rampant parallelism?
 Do
>  non-functional characters give a more consistent signal?  Do we know
>  anything about stratigraphic or geographic separation that might
>  make it extremely unlikely that a given character was synapomorphic
>  in two forms?  None of these approaches may settle a question, but
>  historical sciences (and as a rule, those that use induction)
>  generally rely on what Whewell called consilience: that is, the most
>  robust hypothesis is the one supported by the most lines of
>  independent evidence.  Again, cladograms are simple tools, too
>  simple to encompass this kind of complex historical analysis.  It is
>  already clear that several major questions in dinosaurian phylogeny
>  are dependent on more complex process analyses that have been done
>  so far.  These include the questions of the relationships of
>  marginocephalians to other ornithischians, whether hadrosaurs are
>  diphyletic, and whether certain Triassic Argentinian forms are
>  theropods or even dinosaurs.

An incredibly well-expressed capsule statement about the problems inherent in
cladisitic analyses. Listen up, everyone! This is a master speaking!

G.O.