Paraphyly and names

[SKÁLA Zdenek]

Zdenek Skala:
>> If, e.g., we have a pectinate cladogram of species A,B,C,D,E and
clades
>> A+B and A+B+C+D are supported by many synapomorphies, we will gain
the
>> most informative taxa by grouping ((A+B)+(C+D))+E. This is a
measurable
>> fact ....

Thomas Pape:
>Wrong. Indeed very, very wrong. If the cladogram is pectinate your
analysis provided evidence for (A+B) and for 
>(A + B + C) and for (A + B + C + D). But not for (C + D). With NO
informtion on a 'group' (C + D) it would be 
>scientifically misleading to give this 'taxon' a name.

Read more carefully. Above, I am writing that AB and ABCD are supported
by *many* synapomorphies. Obviously, all clades need to have support to
be clades; this is not the point. Also, bear in mind that each apomorphy
refers to a character state, so there always exists a symmetric
character state (or states) that is (are) plesiomorphic. Hence: if AB is
supported by synapomorphies 1-10, ABC by 11-12, ABCD by 13-23, we can
have taxa (AB and ABCD are without discussion under both alternatives):
(1) either paraphyletic CD (have synapomorphies 13-23 AND
symplesiomorphies 1-10, i.e. is defined by intersection of 2x10
character states)
or
(2) holo(mono)phyletic C and D (each of them having NO autapomorphy,
i.e. no information content from the cladistic point of view) or taxon
ABC (supported by 2 synapomorphies only).
To me, 10 is more than 2, hence the group C+D is superior over the group
A+B+C (not speaking about taxa C and D) in its information content - of
course, unless you say that plesiomorphic character states are not
"real".
Best!
Zdenek Skala
skala at incoma.cz

---
Odchozí zpráva neobsahuje viry.
Zkontrolováno antivirovým systémem AVG (http://www.grisoft.cz).
Verze: 6.0.314 / Virová báze: 175 - datum vydání: 11.1.2002