Re: Gaia theropod follow-up: a "new" phylogeny

[christopher brochu <cbrochu@fieldmuseum.org>]

>
>First of all, I have detected an unspoken attitude among dino-cladists that
>"more is better": "Your analysis only has 258 characters? Hah. >My< analysis
>has 356 characters. Therefore it must be better than yours." But how can you
>compare either analysis against the other? There is simply no metric for
>doing this--certainly none that couldn't be challenged. So--we do a >third<
>analysis, perhaps with >even more< characters. And so on, ad infinitum.


There are several simulation studies out there showing that as we add
information, we more closely approximate the signal supported by a data
set.  Hopefully, this signal is the true phylogeny.  In that sense, larger
matrices genuinely are "better" than smaller ones, provided that character
selection reflects actual variation.




>
>Another problem is that there is no way to assess any one character against
>any other.


In a general sense, there is - a phylogenetic analysis.  But I don't think
that's the point you were making.




 How much is an overhanging zygapophysis worth versus an elongate
>prepubic process?


Depends on the phylogenetic context.  If these are heritable features that
reflect evolutionary history, they both "count" the same.



 As I understand it, most analyses weigh all characters
>equally and hope their numbers will distinguish the apomorphies from the
>homoplasies, but there is no a priori reason for doing the analysis this way,
>and thus no reason to think this weighting will produce the correct
>phylogeny.


Actually, there is an a priori reason - uniform weighting minimizes the
number of ad hoc assumptions added to an analysis.  The old "by assuming
less we learn more" adage.




>My position on morphological cladistic analysis is simply that it is not
>reliable enough to seize the high ground in taxonomy and paleontology the way
>some of the more evangelical cladists think it should: Phylogeny first, then
>everything else. Perfectly correct general principle when the phylogeny is
>tight, but what if the phylogeny is >wrong<? Lots of wasted hypothesizing can
>follow. Cladists seem to decline tests of their phylogenies except via more
>cladistic analysis: If the phylogeny doesn't fit the biogeography or the
>stratigraphy or the functional morphology, well, then the biogeography or the
>stratigraphy or the functional morphology must be wrong or misguided, not the
>phylogeny. But, e.g., when morphological cladograms are tested against
>molecular cladograms, they often do not match. Then what? Which is right--or
>at least, more believable?


Do you actually read any of the papers published in journals like
Systematic Biology, Cladistics, Molecular Phylogenetics and Evolution, or
the like?  You know, the ones filled with a posteriori tests of
phylogenies, many that are not just based on getting a better tree?  I
don't necessarily advocate some of them, but they exist (and they are
certainly used).





chris

_________________________
Christopher A. Brochu
Department of Geology
Field Museum
1400 S. Lake Shore Drive
Chicago IL 60605

312-665-7633 voice
312-665-7641 fax
cbrochu@fieldmuseum.org