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

Re: Classification: A Definition

David Marjanovic writes:
 > Well, this "species and genera" thing is much more complicated.


[Lots snipped since I agree with it all.]

 > Talking about *Mantellisaurus* is useful, talking about
 > *Mantellisaurus atherfieldensis* is nice to Atherfield but
 > otherwise completely useless.

Indeed.  In the dinosaur world, species names are 10% taxonomy, 90% an
opportunity to pay a nice honour to someone you respect :-)

 > Neontological genera, on the other hand, are clades like any other;

... except when they're not, of course.

 > their only function as genera is in Linnaean binominals.

And that is what makes the concept of genus such a problem.  It's why
Cantino et al. (1999) proposed thirteen (thirteen!) methods for
dealing with the problem that genus names are part of species names.

The bottom line, I think, is that there are only two kinds of taxa:
clades, and The Other Kind Of Taxon (hereafter TOKOT).  A TOKOT is a
convenient is defined as "the group of all organisms that are
sufficiently similar to _this_ one" and for the purposes of acting as
a clade specified is precisely equivalent to that one anointed

You _could_ argue that TOKOTs should be abolished completely and all
biology done in terms of specimens and clades only.  But that is never
going to happen and to be honest I wouldn't want it to.  I like the
convenient of TOKOTs.  I prefer the name "Brachiosaurus" to the
specimen number FMNH P 25017.

The problem with Linnaean taxonomy is that instead of just one kind of
TOKOT (species), there are five (species, genus, family, order, class)
or 27 (parvorder, parainfraclass, etc.) or any other number you care
to wish -- except, of course, 1.  And the problem that is inherited by
PN is that the number of different kinds of TOKOT in the Linnaean
system cannot be got below two -- genus and species -- due to the
significance of the former in the naming of the latter.  (At least,
there is no way of getting rid of genera.  It would be theoretically
possible to keep them and to dump species' but I realise that this
idea, which would be abhorent to most dinosaur palaeontologists, would
be simply impossible for angiosperm botanists.)

So now that the PhyloCode has taken the very positive step of
explicitly leaving species-level taxonomy to the rank-based clades,
the challenge is how to get the number of TOKOT types down to one.
That reduced to the problem of uninomial species names, which means
that eight years on we're still right back at Cantino et al. 1999.

Well, not quite -- we have Dayrat's proposal, which looks like an
excellent compromise to me, and is likely to win out.

 > > > > I truly don't know.  It certainly is awfully hard to resist
 > > > > them tempatation to write things like "... and so
 > > > > representatives of at least four sauropod families are known
 > > > > from the Wealden".  And changing that to "at least four
 > > > > clades" only makes things worse.
 > No, it doesn't make things any worse than they already are (as you
 > mentioned earlier in the same post). It only makes them more
 > visible. It takes the disinformation away.

Well; to say that four clades are represented is to say nothing at
all, since each of any four individuals represents a different
sufficiently-small clade from any of the other three.  The question
then becomes whether saying "four families" is not merely low in
information value but negative.  I don't think it is, so I still find
it worth saying, especially in papers that are only likely to be read
be specialists in the particular taxonomic area under discussion, who
will have learned by osmosis how big a "family" is in their area.

 _/|_    ___________________________________________________________________
/o ) \/  Mike Taylor    <mike@indexdata.com>    http://www.miketaylor.org.uk
)_v__/\  "Religion is what you sing on Sunday.  Your true faith is what
         you do all week." -- Parke Godwin.