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

Re: New AVES definition refined (more testable?)



R. Irmis wrote-

> <<However, in truth, this probably does not influence phylogenetic
analyses
> as much as you might think.  Assuming I base my character codings on the
> professionals' illustrations, descriptions and data matrices, I'm no worse
> off than they are.>>
>
> HAHAHAHAHAHA!!!!  Please, tell me another joke, pretty please?  You should
> hear the stories about Gauthier.  When he started his analysis of the
> Saurischia, his initial codings were from literature.  The cladogram that
> resulted was ludicrous...Only when he looked at specimens did he get
> something with decent fidelity.

Okay, if the people who do look at the specimens and publish analyses can't
be trusted to be accurate, then what hope do we have?  I'll spend my time
(which I'd be happy to) and money (which I don't have) visiting other
countries to observe the character states myself and publish my analysis.
Am I supposed to trust my codings then?  You'd think it would be obvious
that my codings are trustable, right?  After all, I SAW the specimens
myself.  Then David Marjanovic wants to do an analysis, and has my data
matrix to base lots of his character codings on.  But will you say it's good
enough for him to do that?  No, you'll say "You can't trust Mickey's
codings, people make mistakes, go spend your time and money looking at the
specimens to verify their accuracy".  What good was all my effort then, if
the data isn't good enough for others to base work on?  Why would David's
observation neccessarily lead to more accurate codings than mine did?  This
is exactly the situation I'm in.  Holtz, Norell and others have already done
the work traveling around and examining specimens, so why should my
observation be any better?  There are controversial codings out there to be
sure, and I think personal verification would be very useful then.  But for
the vast majority of characters, everybody agrees that certain taxa are
coded a certain way.

> <<After all, they based their statements and illustrations on observing
the
> specimen, so I can presumedly trust those codings.  At worst, I'm on par
> with them by perpetuating their errors.>>
>
> Well, you need to learn not to trust them!  People make mistakes,
amazingly
> often.  This is why it is important to look at actual specimens.

Why bother?  You wouldn't trust my observations anyway. ;-)  All I'd be
doing would be verifying codings for my sake, as no one else would be able
to use them without suspicion.  I seriously hope science hasn't come to this
sad state of affairs.

BTW, I'd be VERY surprised if any coelurosaur analysis was based solely on
firsthand examination of specimens.  Through my discussions with various
professionals, it's become clear that they too are mortal, and have limited
cash flows and free time.  Ever notice that the only parts coded in Norell
et al.'s (2001) matrix for Harpymimus are those that are illustrated?
Coincidence?  Look at pg. 1104 of Holtz's breakthrough 1994 analysis.  He
never examined six of his nineteen OTU's.  Rauhut (2000) did very well all
considering, but still has certain taxa that he didn't examine (appendix 1).
So in a perfect world, we would all be millionaires with nothing to do but
reverify previously reverified characters for our own peace of mind, but
that doesn't happen, ever, to anyone.

Mickey Mortimer