Re: Cladospeak (Mammalia, Crurotarsi)

["David Marjanovic" <david.marjanovic@gmx.at>]

----- Original Message -----
From: "Matthew Bonnan" <mbonnan@hotmail.com>
To: <dinosaur@usc.edu>
Sent: Tuesday, September 04, 2001 12:03 AM
Subject: Re: Cladospeak (Mammalia, Crurotarsi)


> Dinogeorge writes:
>
> "I thought groups were >defined< by phyletic relationship, not by
> characters. There are no "group-defining traits" any more."
>
> I'm not sure I follow you.  Please explain.  I am curious to know how a
> phyletic relationship (and thus the definition of a group) can be known
> without reference to specific characters, whether they be synapomorphies
or
> not.  If there are no group-defining traits anymore, how does taxonomy or
> phylogeny work?  Supposedly, one needs group-defining traits in order to
> establish a group?  Are there other ways, apart from divine revelation or
> authoritarian say-so, to establish group identity without using
> group-defining traits?

Such traits are called diagnosing, and they are not part of the definition
(except in apomorphy-based definitions...).

> "Groups may be
>  >diagnosed< by characters, but not defined by them. Unless of course
you're
> still supporting the idea of character-based groups..."

First, a cladistic analysis finds a group because it shares certain
characters. Then someone comes, gives that group a name and defines it --
based on included and excluded species, not on characters (unless s/he
chooses an apomorphy-based definition). The group is then diagnosed by the
traits that unite it; in later analyses the contents and the diagnosis may
change, the definition may not.

Linnean taxa have always had changing contents and diagnoses; the latter
were called definitions but nevertheless subject to change.