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

RE: Classification: A Definition



Christopher Taylor writes:
 > > I have just such an animal in the works at the moment -- clearly
 > > new, very different from anything that already has a name, yet
 > > not resolvable with any confidence at all below the level
 > > Neosauropoda.  I'd call that Neosauropoda incertae sedis; would
 > > you instead prefer to give it a monogeneric family?  If so, to
 > > what end?
 > 
 > Not quite what I meant :-).

Oh.  Well ... good!  :-)

 > However, if you had reasons to believe that your new taxon was
 > probably related to some other neosauropod of equally uncertain
 > position (for the sake of argument, let's say _Euhelopus_) but were
 > unable to rigorously test this phylogenetically, it might still be
 > useful to unite the two as "Euhelopodidae" as a means of
 > highlighting the likely connection. Ditto for if you thought it
 > most likely a member of Diplodocidae but couldn't prove it (maybe
 > the biogeography, age, some other factor that can't easily be
 > entered into a phylogenetic matrix, encourages this idea).
 > [...]
 > the points I've been making are not intended as attacks on
 > phylogenetic nomenclature _per se_. They are arguments about making
 > phylogeny a _sine qua non_ of nomenclature.

Ah, no -- I think what you want to avoid is making _cladistic
analysis_ a sine qua non of nomenclature, not phylogeny.  In the
scenarios you described here, I would be making a perfectly good
phylogenetic hypothesis, and my classification of the new animal would
reflect that phylogenetic hypothesis -- it's just that the phylogeny
would not have come from running PAUP.  Well, no, there's no reason
why it should.

In fact there's an excellent example that we've just been alerted to
today: Kristian Remes's new paper in _Palaeontology_ describing the
new sauropod _Australodocus_.  He did not run a cladistic analysis,
but did use other lines of reasoning to conclude that his new genus is
diplodocine.  On that basis, he is quite happy to assign it to the
formal taxon Diplodocinae -- and quite right, too.  Cladistics is only
one of the tools in the armoury.

 > Of course, this would be a provisional hypothesis only and future
 > researchers may bust this grouping wide open, but isn't that how
 > scientific investigation works?

_All_ hypothesis are provisional!  Otherwise they wouldn't be
hypotheses any more :-)  And hypotheses generated by running a
cladistic analysis are just as provision, just as subject to being
overturned, as those generated in other ways.  (The great advantage
of cladistics is not really that it gives better results than any
other way of hypothesising phylogeny, but that it forces you to "show
your working", so that others can replicate and improve upon it.)

So ... I _think_ we actually agree.  Is it possible?

 _/|_    ___________________________________________________________________
/o ) \/  Mike Taylor    <mike@indexdata.com>    http://www.miketaylor.org.uk
)_v__/\  "They tell me it kind of hurt him bad // Kinda made him feel
         pretty sad" -- Jimi Hendrix waxes poetic on "Highway Chile"