Encroachment and Prohibition (Science and Evolution )

[Ken Kinman]

Richard,
     There are further possibilities to explain your ABC example, and they
are more common and problematic than many strict cladists care to think
about.  Except for polyphyly, probably the worse case scenario for a strict
cladist is that only one of the three is actually holophyletic (C) and the
first two are paraphyletic in serial fashion:
    A ---> B ---> C.    This gives us a total of seven taxa:  three
holophyletic taxa: C and (B + C) and (A + B + C); plus three paraphyletic
taxa: A, B, and (A + B); plus one polyphyletic taxon: (A + C).  So although
the characters shared by B and C are likely synapomorphies, the characters
shared by A and B are likely plesiomorphies, and only those shared by A and
C would be convergences (such as reversals).
     Once we are able to demonstrate such paraphylies (and this can be
extremely difficult), strict cladists "declare" the paraphyletic taxa
"unnatural" and automatically split them.  But the paraphyletic taxon A is
not unnatural just because it is not sister group to (B + C), it just means
that (B + C)'s sister group is a subgroup of A.  That is where my "markers"
come in, and it is thus no longer necessary to automatically split taxon A
EVERY single time it is found to be  paraphyletic.  When you really get down
to brass tacks, the basal taxon almost ALWAYS IS paraphyletic, but we
usually just don't have the fossil evidence to demonstrate it.
     THEREFORE, a cladisto-eclectic classification is actually the most
natural kind of classification.  It is only huge data gaps (often fossils)
which make cladistic classification so useful most of the time.   Paraphyly
is the reality and sister groups a very useful approximation (which makes
cladistics so useful), but at the same time makes strictly cladistic
classifications (across the board) so dangerous, unstable, and impractical.
Thus to use Richard's example, it is the assumption that A, B, (as well as
C) are all really holophyletic (and can truly form sister groups) that is
the often the real problem.  Strict cladism has limits, and pushing it
beyond those limits very quickly becomes counterproductive and can become a
vicious cycle.
               ----- Ken Kinman

*****************************************
>From: Richard Zander <Richard.Zander at MOBOT.ORG>
>Reply-To: Richard Zander <Richard.Zander at MOBOT.ORG>
>To: TAXACOM at USOBI.ORG
>Subject: Re: Encroachment and Prohibition (Science and Evolution )
>Date: Wed, 20 Nov 2002 12:47:41 -0600
>
>Hang on there. Cladistics may be "the best and most objective means of
>reconstructing phylogeny," but there is a big jump from that to "once you
>have reconstructed phylogeny".
>
>The kicker in cladistics is that getting a multifurcation as a result of
>parsimony analysis is the only acceptable time a cladist throws up his/her
>hands and says geeze we can't make a decision. This is nonsense. Suppose we
>have three alternative branches and one of them is supported by only one
>character more than the others? E.g. when closely related branches A, B and
>C are resolved as ((AB)C), but AB is supported by 3 characters and AC by 2
>others and BC by 2 others? (I given this example in previous messages.
>Almost nobody reacted.) Have we established a phylogeny?
>
>If two of the three branches have their support from shared ancestry, then
>the other two got their support randomly from convergence (autapomorphies
>happening to be the same).  Well, support from 3 advanced characters is
>real
>close to support from 2 advanced characters. We haven't established a
>phylogeny.
>
>You need a statistically significant number of characters supporting a
>"best" arrangement of any three branches in a rooted tree against random
>convergence between those three close lineages before you can say you have
>established a phylogeny at some confidence level. This goes for both
>morphological and molecular data.
>
>
>_______________________
>Richard H. Zander
>Missouri Botanical Garden
>PO Box 299
>St. Louis, MO 63166-0299
>Email: richard.zander at mobot.org
>
> > -----Original Message-----
> > From: Steeman, Fedor [SMTP:FSteeman at ZMUC.KU.DK]
> > Sent: Monday, November 18, 2002 8:19 AM
> > To:   TAXACOM at USOBI.ORG
> > Subject:      Re: [TAXACOM] Encroachment and Prohibition (Science and
> > Evolution )
> >
>         ...
> > The cladistic method is the best and most objective one (in a matter of
> > degree) to reconstruct phylogeny with. But once you have reconstructed
> > phylogeny, does classification necessarily have to reflect phylogeny
> > exactly?
>         ...
>
> > The bottom line: you have established a phylogeny, bravo, but let's
>include
> > some extra evolutionary (phenetic) information in the classification. No
> > need for doing things twice.
> >
>         ....
>
> > Kind regards,
> >
> > Fedor Steeman


_________________________________________________________________
MSN 8 helps eliminate e-mail viruses. Get 2 months FREE*.
http://join.msn.com/?page=features/virus